Datasets:

Self-GRIT

/

wikitext-2-raw-v1-preprocessed-5k-PI_KFI_claude-claude-3-5-sonnet-20240620

like 2

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Gods were assumed to be present throughout the world , capable of influencing natural events and the course of human lives . People interacted with them in temples and unofficial shrines , for personal reasons as well as for larger goals of state rites . Egyptians prayed for divine help , used rituals to compel deities to act , and called upon them for advice . Humans ' relations with their gods were a fundamental part of Egyptian society . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n The beings in ancient Egyptian tradition who might be labeled as deities are difficult to count . Egyptian texts list the names of many deities whose nature is unknown and make vague , indirect references to other gods who are not even named . The Egyptologist James P. Allen estimates that more than 1 @,@ 400 deities are named in Egyptian texts , whereas his colleague Christian Leitz says there are \" thousands upon thousands \" of gods . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n The Egyptian language 's terms for these beings were nṯr , \" god \" , and its feminine form nṯrt , \" goddess \" . Scholars have tried to discern the original nature of the gods by proposing etymologies for these words , but none of these suggestions has gained acceptance , and the terms ' origin remains obscure . The hieroglyphs that were used as ideograms and determinatives in writing these words show some of the traits that the Egyptians connected with divinity . The most common of these signs is a flag flying from a pole . Similar objects were placed at the entrances of temples , representing the presence of a deity , throughout ancient Egyptian history . Other such hieroglyphs include a falcon , reminiscent of several early gods who were depicted as falcons , and a seated male or female deity . The feminine form could also be written with an egg as determinative , connecting goddesses with creation and birth , or with a cobra , reflecting the use of the cobra to depict many female deities . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n The Egyptians distinguished nṯrw , \" gods \" , from rmṯ , \" people \" , but the meanings of the Egyptian and the English terms do not match perfectly . The term nṯr may have applied to any being that was in some way outside the sphere of everyday life . Deceased humans were called nṯr because they were considered to be like the gods , whereas the term was rarely applied to many of Egypt 's lesser supernatural beings , which modern scholars often call \" demons \" . Egyptian religious art also depicts places , objects , and concepts in human form . These personified ideas range from deities that were important in myth and ritual to obscure beings , only mentioned once or twice , that may be little more than metaphors . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Confronting these blurred distinctions between gods and other beings , scholars have proposed various definitions of a \" deity \" . One widely accepted definition , suggested by Jan Assmann , says that a deity has a cult , is involved in some aspect of the universe , and is described in mythology or other forms of written tradition . According to a different definition , by Dimitri Meeks , nṯr applied to any being that was the focus of ritual . From this perspective , \" gods \" included the king , who was called a god after his coronation rites , and deceased souls , who entered the divine realm through funeral ceremonies . Likewise , the preeminence of the great gods was maintained by the ritual devotion that was performed for them across Egypt . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n The first written evidence of deities in Egypt comes from the Early Dynastic Period ( c . 3100 – 2686 BC ) . Deities must have emerged sometime in the preceding Predynastic Period ( before 3100 BC ) and grown out of prehistoric religious beliefs . Predynastic artwork depicts a variety of animal and human figures . Some of these images , such as stars and cattle , are reminiscent of important features of Egyptian religion in later times , but in most cases there is not enough evidence to say whether the images are connected with deities . As Egyptian society grew more sophisticated , clearer signs of religious activity appeared . The earliest known temples appeared in the last centuries of the predynastic era , along with images that resemble the iconographies of known deities : the falcon that represents Horus and several other gods , the crossed arrows that stand for Neith , and the enigmatic \" Set animal \" that represents Set . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Many Egyptologists and anthropologists have suggested theories about how the gods developed in these early times . Gustave Jéquier , for instance , thought the Egyptians first revered primitive fetishes , then deities in animal form , and finally deities in human form , whereas Henri Frankfort argued that the gods must have been envisioned in human form from the beginning . Some of these theories are now regarded as too simplistic , and more current ones , such as Siegfried Morenz ' hypothesis that deities emerged as humans began to distinguish themselves from and personify their environment , are difficult to prove . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Predynastic Egypt originally consisted of small , independent villages . Because many deities in later times were strongly tied to particular towns and regions , many scholars have suggested that the pantheon formed as disparate communities coalesced into larger states , spreading and intermingling the worship of the old local deities . But others have argued that the most important predynastic gods were , like other elements of Egyptian culture , present all across the country despite the political divisions within it . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n The final step in the formation of Egyptian religion was the unification of Egypt , in which rulers from Upper Egypt made themselves pharaohs of the entire country . These sacred kings and their subordinates assumed the exclusive right to interact with the gods , and kingship became the unifying focus of the religion . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": "The final step in the formation of Egyptian religion was <|start-PI|> the unification of Egypt , in which <|end-PI|> <|retrieval|> <|start-KFI|> rulers from Upper Egypt made themselves pharaohs of the entire country <|end-KFI|> . <|start-PI|> These sacred kings and their subordinates assumed <|end-PI|> <|retrieval|> <|start-KFI|> the exclusive right to interact with the gods <|end-KFI|> , and <|start-PI|> kingship became the unifying focus of <|end-PI|> <|retrieval|> <|start-KFI|> the religion <|end-KFI|> .", "none": null, "results": [ { "Explanation": "The PI introduces the concept of Egyptian unification as a crucial step in the formation of Egyptian religion, while the KFI specifies who took control during this process. The <|retrieval|> token is inserted to indicate where additional historical context might be needed.", "Ideal Knowledge": "Detailed information about the process of Egyptian unification, including the role of Upper Egyptian rulers in establishing control over the entire country and its impact on religious practices.", "Key Factual Information": "rulers from Upper Egypt made themselves pharaohs of the entire country", "Prefix reformulated": "What happened during the unification of Egypt in relation to Egyptian religion?", "Prefix-intent": "The final step in the formation of Egyptian religion was the unification of Egypt , in which", "Verification": "The KFI directly provides information about what occurred during the unification of Egypt, as prompted by the PI." }, { "Explanation": "The PI sets up a query about the role of sacred kings and their subordinates, while the KFI provides the specific right they assumed. The <|retrieval|> token is inserted to indicate where additional information about religious hierarchy might be needed.", "Ideal Knowledge": "Information on the religious roles and privileges of Egyptian pharaohs and their officials, emphasizing their exclusive status as intermediaries between humans and gods.", "Key Factual Information": "the exclusive right to interact with the gods", "Prefix reformulated": "What did the sacred kings and their subordinates assume in Egyptian religion?", "Prefix-intent": "These sacred kings and their subordinates assumed", "Verification": "The KFI directly states what the sacred kings and their subordinates assumed, as prompted by the PI." }, { "Explanation": "The PI introduces the concept of kingship as a unifying element, while the KFI specifies that it became the focus of the religion. The <|retrieval|> token is inserted to indicate where additional information about the relationship between kingship and religion might be needed.", "Ideal Knowledge": "Explanation of how kingship became central to ancient Egyptian religion, including its role in unifying religious beliefs and practices across the country.", "Key Factual Information": "the religion", "Prefix reformulated": "What did kingship become the unifying focus of in ancient Egypt?", "Prefix-intent": "kingship became the unifying focus of", "Verification": "The KFI directly states what kingship became the unifying focus of, as prompted by the PI." } ] }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n New gods continued to emerge after this transformation . Some important deities like Isis and Amun are not known to have appeared until the Old Kingdom ( c . 2686 – 2181 BC ) . Places and concepts could suddenly inspire the creation of a deity to represent them , and deities were sometimes created to serve as opposite @-@ sex counterparts to established gods or goddesses . Kings were said to be divine , although only a few continued to be worshipped long after their deaths . Some non @-@ royal humans were said to have the favor of the gods and were venerated accordingly . This veneration was usually short @-@ lived , but the court architects Imhotep and Amenhotep son of Hapu were regarded as gods centuries after their lifetimes , as were some other officials . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Through contact with neighboring civilizations , the Egyptians also adopted foreign deities . Dedun , who is first mentioned in the Old Kingdom , may have come from Nubia , and Baal , Anat , and Astarte , among others , were adopted from Canaanite religion during the New Kingdom ( c . 1550 – 1070 BC ) . In Greek and Roman times , from 332 BC to the early centuries AD , deities from across the Mediterranean world were revered in Egypt , but the native gods remained , and they often absorbed the cults of these newcomers into their own worship . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Modern knowledge of Egyptian beliefs about the gods is mostly drawn from religious writings produced by the nation 's scribes and priests . These people were the elite of Egyptian society and were very distinct from the general populace , most of whom were illiterate . Little is known about how well this broader population knew or understood the sophisticated ideas that the elite developed . Commoners ' perceptions of the divine may have differed from those of the priests . The populace may , for example , have mistaken the religion 's symbolic statements about the gods and their actions for literal truth . But overall , what little is known about popular religious belief is consistent with the elite tradition . The two traditions form a largely cohesive vision of the gods and their nature . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Most Egyptian deities represent natural or social phenomena . The gods were generally said to be immanent in these phenomena — to be present within nature . The types of phenomena they represented include physical places and objects as well as abstract concepts and forces . The god Shu was the deification of all the world 's air ; the goddess Meretseger oversaw a limited region of the earth , the Theban Necropolis ; and the god Sia personified the abstract notion of perception . Major gods often had many roles and were involved in several types of phenomena . For instance , Khnum was the god of Elephantine Island in the midst of the Nile , the river that was essential to Egyptian civilization . He was credited with producing the annual Nile flood that fertilized the nation 's farmland . Perhaps as an outgrowth of this life @-@ giving function , he was said to create all living things , fashioning their bodies on a potter 's wheel . Gods could share the same role in nature ; Ra , Atum , Khepri , Horus , and other deities acted as sun gods . Despite their diverse functions , most gods had an overarching role in common : maintaining maat , the universal order that was a central principle of Egyptian religion and was itself personified as a goddess . But some deities represented disruption to maat . Most prominently , Apep was the force of chaos , constantly threatening to annihilate the order of the universe , and Set was an ambivalent member of divine society who could both fight disorder and foment it . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Not all aspects of existence were seen as deities . Although many deities were connected with the Nile , no god personified it in the way that Ra personified the sun . Short @-@ lived phenomena , like rainbows or eclipses , were not represented by gods ; neither were elements like fire and water or many other components of the world . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n The roles of each deity were fluid , and each god could expand its nature to take on new characteristics . As a result , gods ' roles are difficult to categorize or define . But despite their flexibility , the gods had limited abilities and spheres of influence . Not even the creator god could reach beyond the boundaries of the cosmos that he created , and even Isis , though she was said to be the cleverest of the gods , was not omniscient . Richard H. Wilkinson , however , argues that some texts from the late New Kingdom suggest that , as beliefs about the god Amun evolved , he was thought to approach omniscience and omnipresence and to transcend the limits of the world in a way that other deities did not . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n The deities with the most limited and specialized domains are often called \" minor divinities \" or \" demons \" in modern writing , although there is no firm definition for these terms . Among these lesser deities , Egyptologist Claude Traunecker draws a distinction between \" genies \" — specialized patron spirits of certain places , objects , or activities , such as the sea or marsh god Wadj @-@ Wer and the harvest goddess Renenutet — and demons , who have a more dangerous character . Many demons are hostile , causing illness and other troubles among humans . Their power can also be protective ; they may guard certain places in the Duat , the realm of the dead , or advise and watch over humans . Egyptians believed the landscape was full of these unpredictable divine powers . Demons often act as servants and messengers to the greater gods , but their position in the hierarchy is not fixed . The protective deities Bes and Taweret originally had minor , demon @-@ like roles , but over time they came to be credited with great influence . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Divine behavior was believed to govern all of nature . Except for the few deities who disrupted the divine order , the gods ' actions maintained maat and created and sustained all living things . They did this work using a force the Egyptians called heka , a term usually translated as \" magic \" . Heka was a fundamental power that the creator god used to form the world and the gods themselves . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": "<|start-PI|>Divine behavior was believed to govern<|end-PI|> <|retrieval|> <|start-KFI|>all of nature<|end-KFI|>. <|start-PI|>Except for the few deities who disrupted the divine order, the gods' actions maintained<|end-PI|> <|retrieval|> <|start-KFI|>maat and created and sustained all living things<|end-KFI|>. <|start-PI|>They did this work using a force the Egyptians called<|end-PI|> <|retrieval|> <|start-KFI|>heka, a term usually translated as \"magic\"<|end-KFI|>. <|start-PI|>Heka was a fundamental power that the creator god used to<|end-PI|> <|retrieval|> <|start-KFI|>form the world and the gods themselves<|end-KFI|>.", "none": null, "results": [ { "Explanation": "The PI introduces the concept of divine behavior's influence, while the KFI specifies its scope. The <|retrieval|> token is inserted to indicate where additional information about ancient Egyptian beliefs might be needed.", "Ideal Knowledge": "Information on ancient Egyptian beliefs about the role of gods in nature, emphasizing their control over natural phenomena and the world order.", "Key Factual Information": "all of nature", "Prefix reformulated": "What did ancient Egyptians believe divine behavior governed?", "Prefix-intent": "Divine behavior was believed to govern", "Verification": "The KFI directly answers what divine behavior was believed to govern, as prompted by the PI." }, { "Explanation": "The PI sets up the context of gods' actions, and the KFI provides the specific concept they maintained (maat) and their role in creation. The <|retrieval|> token is inserted to indicate where additional information about maat and the gods' roles might be needed.", "Ideal Knowledge": "Detailed explanation of the concept of maat in ancient Egyptian religion, its importance in maintaining cosmic order, and the gods' role in sustaining life.", "Key Factual Information": "maat and created and sustained all living things", "Prefix reformulated": "What did the gods' actions maintain in ancient Egyptian belief?", "Prefix-intent": "Except for the few deities who disrupted the divine order, the gods' actions maintained", "Verification": "The KFI directly states what the gods' actions maintained and created, as prompted by the PI." }, { "Explanation": "The PI introduces the idea of a force used by the gods, and the KFI provides its name and common translation. The <|retrieval|> token is inserted to indicate where additional information about heka might be needed.", "Ideal Knowledge": "Information on the concept of heka in ancient Egyptian religion, its significance, and how it was believed to be used by the gods in maintaining the universe.", "Key Factual Information": "heka, a term usually translated as \"magic\"", "Prefix reformulated": "What force did ancient Egyptians believe gods used in their work?", "Prefix-intent": "They did this work using a force the Egyptians called", "Verification": "The KFI directly provides the name and translation of the force used by the gods, as prompted by the PI." }, { "Explanation": "The PI sets up the context of heka as a fundamental power, and the KFI specifies its use in creation. The <|retrieval|> token is inserted to indicate where additional information about the creator god and the use of heka in creation might be needed.", "Ideal Knowledge": "Detailed explanation of the creation myths in ancient Egyptian religion, focusing on the role of heka in the formation of the world and the gods.", "Key Factual Information": "form the world and the gods themselves", "Prefix reformulated": "What did ancient Egyptians believe the creator god used heka for?", "Prefix-intent": "Heka was a fundamental power that the creator god used to", "Verification": "The KFI directly states what heka was used for by the creator god, as prompted by the PI." } ] }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n The gods ' actions in the present are described and praised in hymns and funerary texts . In contrast , mythology mainly concerns the gods ' actions during a vaguely imagined past in which the gods were present on earth and interacted directly with humans . The events of this past time set the pattern for the events of the present . Periodic occurrences were tied to events in the mythic past ; the succession of each new pharaoh , for instance , reenacted Horus ' accession to the throne of his father Osiris . Myths are metaphors for the gods ' actions , which humans cannot fully understand . They contain seemingly contradictory ideas , each expressing a particular perspective on divine events . The contradictions in myth are part of the Egyptians ' many @-@ faceted approach to religious belief — what Henri Frankfort called a \" multiplicity of approaches \" to understanding the gods . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n In myth , the gods behave much like humans . They feel emotion ; they can eat , drink , fight , weep , sicken , and die . Some have unique character traits . Set is aggressive and impulsive , and Thoth , patron of writing and knowledge , is prone to long @-@ winded speeches . Yet overall , the gods are more like archetypes than well drawn characters . Their behavior is inconsistent , and their thoughts and motivations are rarely stated . Most myths about them lack highly developed characters and plots , because the symbolic meaning of the myths was more important than elaborate storytelling . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n The first divine act is the creation of the cosmos , described in several creation myths . They focus on different gods , each of which may act as creator deities . The eight gods of the Ogdoad , who represent the chaos that precedes creation , give birth to the sun god , who establishes order in the newly formed world ; Ptah , who embodies thought and creativity , gives form to all things by envisioning and naming them ; Atum produces all things as emanations of himself ; and Amun , according to the myths promoted by his priesthood , preceded and created the other creator gods . These and other versions of the events of creation were not seen as contradictory . Each gives a different perspective on the complex process by which the organized universe and its many deities emerged from undifferentiated chaos . The period following creation , in which a series of gods rule as kings over the divine society , is the setting for most myths . The gods struggle against the forces of chaos and among each other before withdrawing from the human world and installing the historical kings of Egypt to rule in their place . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n A recurring theme in these myths is the effort of the gods to maintain maat against the forces of disorder . They fight vicious battles with the forces of chaos at the start of creation . Ra and Apep , battling each other each night , continue this struggle into the present . Another prominent theme is the gods ' death and revival . The clearest instance where a god dies is the myth of Osiris ' murder , in which that god is resurrected as ruler of the Duat . The sun god is also said to grow old during his daily journey across the sky , sink into the Duat at night , and emerge as a young child at dawn . In the process he comes into contact with the rejuvenating water of primordial chaos . Funerary texts that depict Ra 's journey through the Duat also show the corpses of gods who are enlivened along with him . Instead of being changelessly immortal , the gods periodically died and were reborn by repeating the events of creation , thus renewing the whole world . But it was always possible for this cycle to be disrupted and for chaos to return . Some poorly understood Egyptian texts even suggest that this calamity is destined to happen — that the creator god will one day dissolve the order of the world , leaving only himself and Osiris amid the primordial chaos . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Gods were linked with specific regions of the universe . In Egyptian tradition , the world includes the earth , the sky , and the Duat . Surrounding them is the dark formlessness that existed before creation . The gods in general were said to dwell in the sky , although gods whose roles were linked with other parts of the universe were said to live in those places instead . Most events of mythology , set in a time before the gods ' withdrawal from the human realm , take place in an earthly setting . The deities there sometimes interact with those in the sky . The Duat , in contrast , is treated as a remote and inaccessible place , and the gods who dwell there have difficulty communicating with those in the world of the living . The space outside the cosmos is also said to be very distant . It too is inhabited by deities , some hostile and some beneficial to the other gods and their orderly world . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n In the time after myth , most gods were said to be either in the sky or invisibly present within the world . Temples were their main means of contact with humanity . Each day , it was believed , the gods moved from the divine realm to their temples , their homes in the human world . There they inhabited the cult images , the statues that depicted deities and allowed humans to interact with them in temple rituals . This movement between realms was sometimes described as a journey between the sky and the earth . As temples were the focal points of Egyptian cities , the god in a city 's main temple was the patron god for the city and the surrounding region . Deities ' spheres of influence on earth centered on the towns and regions they presided over . Many gods had more than one cult center , and their local ties changed over time . They could establish themselves in new cities , or their range of influence could contract . Therefore , a given deity 's main cult center in historical times is not necessarily his or her place of origin . The political influence of a city could affect the importance of its patron deity . When kings from Thebes took control of the country at start of the Middle Kingdom ( c . 2055 – 1650 BC ) , they elevated Thebes ' patron gods — first the war god Montu and then Amun — to national prominence . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n In Egyptian belief , names express the fundamental nature of the things to which they refer . In keeping with this belief , the names of deities often relate to their roles or origins . The name of the predatory goddess Sekhmet means \" powerful one \" , the name of the mysterious god Amun means \" hidden one \" , and the name of the goddess Nekhbet , who was worshipped in the city of Nekheb , means \" she of Nekheb \" . But many other names have no certain meaning , even when the gods who bear them are closely tied to a single role . The names of the sky goddess Nut and the earth god Geb do not resemble the Egyptian terms for sky and earth . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n The Egyptians also devised false etymologies giving more meanings to divine names . A passage in the Coffin Texts renders the name of the funerary god Sokar as sk r , meaning \" cleaning of the mouth \" , to link his name with his role in the Opening of the Mouth ritual , while one in the Pyramid Texts says the name is based on words shouted by Osiris , connecting Sokar with the most important funerary deity . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n The gods were believed to have many names . Among them were secret names that conveyed their true natures more profoundly than others . To know the true name of a deity was to have power over it . The importance of names is demonstrated by a myth in which Isis poisons the superior god Ra and refuses to cure him unless he reveals his secret name to her . Upon learning the name , she tells it to her son , Horus , and by learning it they gain greater knowledge and power . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n In addition to their names , gods were given epithets , like \" possessor of splendor \" , \" ruler of Abydos \" , or \" lord of the sky \" , that describe some aspect of their roles or their worship . Because of the gods ' multiple and overlapping roles , deities can have many epithets — with more important gods accumulating more titles — and the same epithet can apply to many deities . Some epithets eventually became separate deities , as with Werethekau , an epithet applied to several goddesses meaning \" great enchantress \" , which came to be treated as an independent goddess . The host of divine names and titles expresses the gods ' multifarious nature . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Egyptian deities are connected in a complex and shifting array of relationships . A god 's connections and interactions with other deities helped define its character . Thus Isis , as the mother and protector of Horus , was a great healer as well as the patroness of kings . Such relationships were the base material from which myths were formed . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Family relationships are a common type of connection between gods . Deities often form male and female pairs , reflecting the importance of procreation in Egyptian religious thought . Families of three deities , with a father , mother , and child , represent the creation of new life and the succession of the father by the child , a pattern that connects divine families with royal succession . Osiris , Isis , and Horus formed the quintessential family of this type . The pattern they set grew more widespread over time , so that many deities in local cult centers , like Ptah , Sekhmet , and their child Nefertum at Memphis and Amun , Mut , and Khonsu at Thebes , were assembled into family triads . Genealogical connections like these are changeable , in keeping with the multiple perspectives in Egyptian belief . Hathor , as a fertility goddess , could act as mother to any child god , including the child form of the sun god , although in other circumstances she was the sun god 's daughter . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Other divine groups were composed of deities with interrelated roles , or who together represented a region of the Egyptian mythological cosmos . There were sets of gods for the hours of the day and night and for each nome ( province ) of Egypt . Some of these groups contain a specific , symbolically important number of deities . Paired gods can stand for opposite but interrelated concepts that are part of a greater unity . Ra , who is dynamic and light @-@ producing , and Osiris , who is static and shrouded in darkness , merge into a single god each night . Groups of three are linked with plurality in ancient Egyptian thought , and groups of four connote completeness . Rulers in the late New Kingdom promoted a particularly important group of three gods above all others : Amun , Ra , and Ptah . These deities stood for the plurality of all gods , as well as for their own cult centers ( the major cities of Thebes , Heliopolis , and Memphis ) and for many threefold sets of concepts in Egyptian religious thought . Sometimes Set , the patron god of the Nineteenth Dynasty kings and the embodiment of disorder within the world , was added to this group , which emphasized a single coherent vision of the pantheon . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Nine , the product of three and three , represents a multitude , so the Egyptians called several large groups \" enneads \" , or sets of nine , even if they had more than nine members . The most prominent ennead was the Ennead of Heliopolis , an extended family of deities descended from the creator god Atum , which incorporates many important gods . The term \" ennead \" was often extended to include all of Egypt 's deities . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": "<|start-PI|>Nine , the product of three and three , represents<|end-PI|> <|retrieval|> <|start-KFI|>a multitude<|end-KFI|> , so <|start-PI|>The Egyptians called several large groups<|end-PI|> <|retrieval|> <|start-KFI|>\" enneads \" , or sets of nine , even if they had more than nine members<|end-KFI|> . <|start-PI|>The most prominent ennead was<|end-PI|> <|retrieval|> <|start-KFI|>the Ennead of Heliopolis , an extended family of deities descended from the creator god Atum , which incorporates many important gods<|end-KFI|> . <|start-PI|>The term \" ennead \" was often extended to include<|end-PI|> <|retrieval|> <|start-KFI|>all of Egypt 's deities<|end-KFI|> .", "none": null, "results": [ { "Explanation": "The PI introduces the number nine and sets up a query about its representation, while the KFI provides the specific meaning. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.", "Ideal Knowledge": "Information about the symbolic and cultural significance of the number nine in various civilizations, particularly in ancient Egypt.", "Key Factual Information": "a multitude", "Prefix reformulated": "What does the number nine represent?", "Prefix-intent": "Nine , the product of three and three , represents", "Verification": "The KFI directly provides the representation of the number nine as mentioned in the PI." }, { "Explanation": "The PI sets up a query about Egyptian terminology for large groups, and the KFI provides the specific term and its meaning. The <|retrieval|> token is inserted to indicate where additional information about Egyptian cultural concepts might be needed.", "Ideal Knowledge": "Detailed explanation of the Egyptian concept of 'enneads', including its etymology, usage in various contexts, and examples of specific enneads in Egyptian mythology.", "Key Factual Information": "\" enneads \" , or sets of nine , even if they had more than nine members", "Prefix reformulated": "What did the Egyptians call several large groups?", "Prefix-intent": "The Egyptians called several large groups", "Verification": "The KFI directly provides the term used by Egyptians for large groups, as prompted by the PI." }, { "Explanation": "The PI introduces the concept of a prominent ennead, and the KFI provides specific details about the Ennead of Heliopolis. The <|retrieval|> token is inserted to indicate where additional information about Egyptian mythology might be needed.", "Ideal Knowledge": "Comprehensive information about the Ennead of Heliopolis, including its composition, the roles of individual deities, and its significance in Egyptian religion and mythology.", "Key Factual Information": "the Ennead of Heliopolis , an extended family of deities descended from the creator god Atum , which incorporates many important gods", "Prefix reformulated": "What was the most prominent ennead in Egyptian mythology?", "Prefix-intent": "The most prominent ennead was", "Verification": "The KFI directly provides information about the most prominent ennead, as prompted by the PI." }, { "Explanation": "The PI sets up a query about the extended use of the term 'ennead', and the KFI provides the specific scope of its application. The <|retrieval|> token is inserted to indicate where additional information about the evolution of Egyptian religious terminology might be needed.", "Ideal Knowledge": "Explanation of how the term 'ennead' evolved in Egyptian mythology to encompass a broader range of deities beyond its original meaning, including examples of how it was applied to various groupings of gods.", "Key Factual Information": "all of Egypt 's deities", "Prefix reformulated": "What did the term 'ennead' often include in Egyptian mythology?", "Prefix-intent": "The term \" ennead \" was often extended to include", "Verification": "The KFI directly provides information about the extended use of the term 'ennead', as prompted by the PI." } ] }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n This divine assemblage had a vague and changeable hierarchy . Gods with broad influence in the cosmos or who were mythologically older than others had higher positions in divine society . At the apex of this society was the king of the gods , who was usually identified with the creator deity . In different periods of Egyptian history , different gods were most frequently said to hold this exalted position . Horus was the most important god in the Early Dynastic Period , Ra rose to preeminence in the Old Kingdom , Amun was supreme in the New , and in the Ptolemaic and Roman periods , Isis was the divine queen and creator goddess . Newly prominent gods tended to adopt characteristics from their predecessors . Isis absorbed the traits of many other goddesses during her rise , and when Amun became the ruler of the pantheon , he was conjoined with Ra , the traditional king of the gods , to become a solar deity . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n The gods were believed to manifest in many forms . The Egyptians had complex conception of the human soul , consisting of several parts . The spirits of the gods were composed of many of these same elements . The ba was the component of the human or divine soul that affected the world around it . Any visible manifestation of a god 's power could be called its ba ; thus , the sun was called the ba of Ra . A depiction of a deity was considered a ka , another component of its being , which acted as a vessel for that deity 's ba to inhabit . The cult images of gods that were the focus of temple rituals , as well as the sacred animals that represented certain deities , were believed to house divine bas in this way . Gods could be ascribed many bas and kas , which were sometimes given names representing different aspects of the god 's nature . Everything in existence was said to be one of the kas of Atum the creator god , who originally contained all things within himself , and one deity could be called the ba of another , meaning that the first god is a manifestation of the other 's power . Divine body parts could act as separate deities , like the Eye of Ra and Hand of Atum , both of which were personified as goddesses . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": "The gods were believed to manifest in many forms . <|start-PI|> The Egyptians had complex conception of the human soul , consisting of <|end-PI|> <|retrieval|> <|start-KFI|> several parts <|end-KFI|> . The spirits of the gods were composed of many of these same elements . <|start-PI|> The ba was the component of the human or divine soul that <|end-PI|> <|retrieval|> <|start-KFI|> affected the world around it <|end-KFI|> . Any visible manifestation of a god 's power could be called its ba ; thus , the sun was called the ba of Ra . <|start-PI|> A depiction of a deity was considered a ka , another component of its being , which <|end-PI|> <|retrieval|> <|start-KFI|> acted as a vessel for that deity 's ba to inhabit <|end-KFI|> . The cult images of gods that were the focus of temple rituals , as well as the sacred animals that represented certain deities , were believed to house divine bas in this way . Gods could be ascribed many bas and kas , which were sometimes given names representing different aspects of the god 's nature . <|start-PI|> Everything in existence was said to be one of the kas of <|end-PI|> <|retrieval|> <|start-KFI|> Atum the creator god , who originally contained all things within himself <|end-KFI|> , and one deity could be called the ba of another , meaning that the first god is a manifestation of the other 's power . Divine body parts could act as separate deities , like the Eye of Ra and Hand of Atum , both of which were personified as goddesses .", "none": null, "results": [ { "Explanation": "The PI introduces the Egyptian concept of the soul, and the KFI provides the key information that it consists of multiple parts. The <|retrieval|> token is inserted to indicate where additional information about these parts might be needed.", "Ideal Knowledge": "A detailed explanation of the ancient Egyptian concept of the soul, including its various components such as the ba, ka, and other elements.", "Key Factual Information": "several parts", "Prefix reformulated": "What did the ancient Egyptians believe the human soul consisted of?", "Prefix-intent": "The Egyptians had complex conception of the human soul , consisting of", "Verification": "The KFI directly answers the question posed in the PI about the composition of the human soul according to Egyptian belief." }, { "Explanation": "The PI introduces the ba as a component of the soul, and the KFI provides its specific function. The <|retrieval|> token is inserted to indicate where additional information about the ba's role might be needed.", "Ideal Knowledge": "A comprehensive explanation of the ba in ancient Egyptian religion, including its role in affecting the physical world and its manifestations in various forms.", "Key Factual Information": "affected the world around it", "Prefix reformulated": "What was the function of the ba in ancient Egyptian belief?", "Prefix-intent": "The ba was the component of the human or divine soul that", "Verification": "The KFI directly describes the function of the ba as stated in the PI." }, { "Explanation": "The PI introduces the ka as another component of a deity's being, and the KFI provides its specific function. The <|retrieval|> token is inserted to indicate where additional information about the ka's role might be needed.", "Ideal Knowledge": "An in-depth explanation of the ka in ancient Egyptian religion, including its relationship to physical representations of deities and its role in housing the divine ba.", "Key Factual Information": "acted as a vessel for that deity 's ba to inhabit", "Prefix reformulated": "What was the function of a ka in ancient Egyptian belief?", "Prefix-intent": "A depiction of a deity was considered a ka , another component of its being , which", "Verification": "The KFI directly explains the function of the ka as described in the PI." }, { "Explanation": "The PI sets up the concept of everything being a ka, and the KFI identifies Atum as the source. The <|retrieval|> token is inserted to indicate where additional information about Atum and his role might be needed.", "Ideal Knowledge": "A detailed account of Atum in ancient Egyptian mythology, including his role as the creator god and the concept of all existence being derived from his kas.", "Key Factual Information": "Atum the creator god , who originally contained all things within himself", "Prefix reformulated": "According to ancient Egyptian belief, whose kas did everything in existence represent?", "Prefix-intent": "Everything in existence was said to be one of the kas of", "Verification": "The KFI directly answers the question posed in the PI, identifying Atum as the source of all kas." } ] }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Nationally important deities gave rise to local manifestations , which sometimes absorbed the characteristics of older regional gods . Horus had many forms tied to particular places , including Horus of Nekhen , Horus of Buhen , and Horus of Edfu . Such local manifestations could be treated almost as separate beings . During the New Kingdom , one man was accused of stealing clothes by an oracle supposed to communicate messages from Amun of Pe @-@ Khenty . He consulted two other local oracles of Amun hoping for a different judgment . Gods ' manifestations also differed according to their roles . Horus could be a powerful sky god or vulnerable child , and these forms were sometimes counted as independent deities . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Gods were combined with each other as easily as they were divided . A god could be called the ba of another , or two or more deities could be joined into one god with a combined name and iconography . Local gods were linked with greater ones , and deities with similar functions were combined . Ra was connected with the local deity Sobek to form Sobek @-@ Ra ; with his fellow ruling god , Amun , to form Amun @-@ Ra ; with the solar form of Horus to form Ra @-@ Horakhty ; and with several solar deities as Horemakhet @-@ Khepri @-@ Ra @-@ Atum . On rare occasion , even deities of different sexes were joined in this way , producing combinations like Osiris @-@ Neith and Mut @-@ Min . This linking of deities is called syncretism . Unlike other situations for which this term is used , the Egyptian practice was not meant to fuse competing belief systems , although foreign deities could be syncretized with native ones . Instead , syncretism acknowledged the overlap between their roles , and extended the sphere of influence for each of them . Syncretic combinations were not permanent ; a god who was involved in one combination continued to appear separately and to form new combinations with other deities . But closely connected deities did sometimes merge . Horus absorbed several falcon gods from various regions , such as Khenty @-@ irty and Khenty @-@ khety , who became little more than local manifestations of him ; Hathor subsumed a similar cow goddess , Bat ; and an early funerary god , Khenti @-@ Amentiu , was supplanted by Osiris and Anubis . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n In the reign of Akhenaten ( c . 1353 – 1336 BC ) in the mid @-@ New Kingdom , a single solar deity , the Aten , became the sole focus of the state religion . Akhenaten ceased to fund the temples of other deities and erased the gods ' names and images on monuments , targeting Amun in particular . This new religious system , sometimes called Atenism , differed dramatically from the polytheistic worship of many gods in all other periods . Whereas , in earlier times , newly important gods were integrated into existing religious beliefs , Atenism insisted on a single understanding of the divine that excluded the traditional multiplicity of perspectives . Yet Atenism may not have been full monotheism , which totally excludes belief in other deities . There is evidence suggesting that the general populace was still allowed to worship other gods in private . The picture is further complicated by Atenism 's apparent tolerance for some other deities , like Shu . For these reasons , the Egyptologist Dominic Montserrat suggested that Akhenaten may have been monolatrous , worshipping a single deity while acknowledging the existence of others . In any case , Atenism 's aberrant theology did not take root among the Egyptian populace , and Akhenaten 's successors returned to traditional beliefs . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Scholars have long debated whether traditional Egyptian religion ever asserted that the multiple gods were , on a deeper level , unified . Reasons for this debate include the practice of syncretism , which might suggest that all the separate gods could ultimately merge into one , and the tendency of Egyptian texts to credit a particular god with power that surpasses all other deities . Another point of contention is the appearance of the word \" god \" in wisdom literature , where the term does not refer to a specific deity or group of deities . In the early 20th century , for instance , E. A. Wallis Budge believed that Egyptian commoners were polytheistic , but knowledge of the true monotheistic nature of the religion was reserved for the elite , who wrote the wisdom literature . His contemporary James Henry Breasted thought Egyptian religion was instead pantheistic , with the power of the sun god present in all other gods , while Hermann Junker argued that Egyptian civilization had been originally monotheistic and became polytheistic in the course of its history . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n In 1971 , Erik Hornung published a study rebutting these views . He points out that in any given period many deities , even minor ones , were described as superior to all others . He also argues that the unspecified \" god \" in the wisdom texts is a generic term for whichever deity the reader chooses to revere . Although the combinations , manifestations , and iconographies of each god were constantly shifting , they were always restricted to a finite number of forms , never becoming fully interchangeable in a monotheistic or pantheistic way . Henotheism , Hornung says , describes Egyptian religion better than other labels . An Egyptian could worship any deity at a particular time and credit it with supreme power in that moment , without denying the other gods or merging them all with the god that he or she focused on . Hornung concludes that the gods were fully unified only in myth , at the time before creation , after which the multitude of gods emerged from a uniform nonexistence . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Hornung 's arguments have greatly influenced other scholars of Egyptian religion , but some still believe that at times the gods were more unified than he allows . Jan Assmann maintains that the notion of a single deity developed slowly through the New Kingdom , beginning with a focus on Amun @-@ Ra as the all @-@ important sun god . In his view , Atenism was an extreme outgrowth of this trend . It equated the single deity with the sun and dismissed all other gods . Then , in the backlash against Atenism , priestly theologians described the universal god in a different way , one that coexisted with traditional polytheism . The one god was believed to transcend the world and all the other deities , while at the same time , the multiple gods were aspects of the one . According to Assmann , this one god was especially equated with Amun , the dominant god in the late New Kingdom , whereas for the rest of Egyptian history the universal deity could be identified with many other gods . James P. Allen says that coexisting notions of one god and many gods would fit well with the \" multiplicity of approaches \" in Egyptian thought , as well as with the henotheistic practice of ordinary worshippers . He says that the Egyptians may have recognized the unity of the divine by \" identifying their uniform notion of ' god ' with a particular god , depending on the particular situation . \" \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Egyptian writings describe the gods ' bodies in detail . They are made of precious materials ; their flesh is gold , their bones are silver , and their hair is lapis lazuli . They give off a scent that the Egyptians likened to the incense used in rituals . Some texts give precise descriptions of particular deities , including their height and eye color . Yet these characteristics are not fixed ; in myths , gods change their appearances to suit their own purposes . Egyptian texts often refer to deities ' true , underlying forms as \" mysterious \" . The Egyptians ' visual representations of their gods are therefore not literal . They symbolize specific aspects of each deity 's character , functioning much like the ideograms in hieroglyphic writing . For this reason , the funerary god Anubis is commonly shown in Egyptian art as a dog or jackal , a creature whose scavenging habits threaten the preservation of buried mummies , in an effort to counter this threat and employ it for protection . His black coloring alludes to the color of mummified flesh and to the fertile black soil that Egyptians saw as a symbol of resurrection . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Most gods were depicted in several ways . Hathor could be a cow , cobra , lioness , or a woman with bovine horns or ears . By depicting a given god in different ways , the Egyptians expressed different aspects of its essential nature . The gods are depicted in a finite number of these symbolic forms , so that deities can often be distinguished from one another by their iconographies . These forms include men and women ( anthropomorphism ) , animals ( zoomorphism ) , and , more rarely , inanimate objects . Combinations of forms , such as gods with human bodies and animal heads , are common . New forms and increasingly complex combinations arose in the course of history . Some gods can only be distinguished from others if they are labeled in writing , as with Isis and Hathor . Because of the close connection between these goddesses , they could both wear the cow @-@ horn headdress that was originally Hathor 's alone . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Certain features of divine images are more useful than others in determining a god 's identity . The head of a given divine image is particularly significant . In a hybrid image , the head represents the original form of the being depicted , so that , as the Egyptologist Henry Fischer put it , \" a lion @-@ headed goddess is a lion @-@ goddess in human form , while a royal sphinx , conversely , is a man who has assumed the form of a lion . \" Divine headdresses , which range from the same types of crowns used by human kings to large hieroglyphs worn on gods ' heads , are another important indicator . In contrast , the objects held in gods ' hands tend to be generic . Male deities hold was staffs , goddesses hold stalks of papyrus , and both sexes carry ankh signs , representing the Egyptian word for \" life \" , to symbolize their life @-@ giving power . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n The forms in which the gods are shown , although diverse , are limited in many ways . Many creatures that are widespread in Egypt were never used in divine iconography , whereas a few , such as falcons , cobras , and cattle , can each represent many deities . Animals that were absent from Egypt in the early stages of its history were not used as divine images . For instance , the horse , which was only introduced in the Second Intermediate Period ( c . 1650 – 1550 BC ) , never represented a god . Similarly , the clothes worn by anthropomorphic deities in all periods changed little from the styles used in the Old Kingdom : a kilt , false beard , and often a shirt for male gods and a long , tight @-@ fitting dress for goddesses . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": "<|start-PI|>The forms in which the gods are shown , although diverse , are limited in many ways<|end-PI|> <|retrieval|> <|start-KFI|>Many creatures that are widespread in Egypt were never used in divine iconography , whereas a few , such as falcons , cobras , and cattle , can each represent many deities<|end-KFI|> . <|start-PI|>Animals that were absent from Egypt in the early stages of its history were not used as divine images<|end-PI|> <|retrieval|> <|start-KFI|>For instance , the horse , which was only introduced in the Second Intermediate Period ( c . 1650 – 1550 BC ) , never represented a god<|end-KFI|> . <|start-PI|>Similarly , the clothes worn by anthropomorphic deities in all periods changed little from the styles used in the Old Kingdom<|end-PI|> <|retrieval|> <|start-KFI|>a kilt , false beard , and often a shirt for male gods and a long , tight @-@ fitting dress for goddesses<|end-KFI|> .", "none": null, "results": [ { "Explanation": "The PI introduces the concept of limitations in divine forms, while the KFI provides specific examples of these limitations. The <|retrieval|> token is inserted to indicate where additional information about Egyptian divine iconography might be needed.", "Ideal Knowledge": "A comprehensive list of animals and creatures used in Egyptian divine iconography, including their frequency of use and the deities they represent. This should also include information on which common Egyptian animals were not used in divine representations.", "Key Factual Information": "Many creatures that are widespread in Egypt were never used in divine iconography , whereas a few , such as falcons , cobras , and cattle , can each represent many deities", "Prefix reformulated": "How are the forms of Egyptian gods limited?", "Prefix-intent": "The forms in which the gods are shown , although diverse , are limited in many ways", "Verification": "The KFI directly provides examples of limitations in divine iconography, as prompted by the PI." }, { "Explanation": "The PI sets up a rule about animals not present in early Egyptian history, and the KFI provides a specific example with the horse. The <|retrieval|> token is inserted to indicate where additional information about animal introductions and their use in iconography might be needed.", "Ideal Knowledge": "A timeline of animal introductions to Egypt, focusing on those that came after the establishment of the pantheon, with specific information on the horse's introduction during the Second Intermediate Period and its absence from divine representations.", "Key Factual Information": "For instance , the horse , which was only introduced in the Second Intermediate Period ( c . 1650 – 1550 BC ) , never represented a god", "Prefix reformulated": "What is an example of an animal not used as a divine image in ancient Egypt?", "Prefix-intent": "Animals that were absent from Egypt in the early stages of its history were not used as divine images", "Verification": "The KFI directly provides an example of an animal not used in divine imagery, as prompted by the PI." }, { "Explanation": "The PI introduces the concept of consistent clothing styles for deities, and the KFI provides specific examples for both male and female gods. The <|retrieval|> token is inserted to indicate where additional information about divine clothing styles might be needed.", "Ideal Knowledge": "Detailed descriptions and visual representations of clothing styles for Egyptian deities from the Old Kingdom through later periods, highlighting the consistency in fashion and any minor variations over time.", "Key Factual Information": "a kilt , false beard , and often a shirt for male gods and a long , tight @-@ fitting dress for goddesses", "Prefix reformulated": "What were the typical clothes worn by anthropomorphic deities in ancient Egypt?", "Prefix-intent": "Similarly , the clothes worn by anthropomorphic deities in all periods changed little from the styles used in the Old Kingdom", "Verification": "The KFI directly describes the typical clothing for male and female deities, as prompted by the PI." } ] }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n The basic anthropomorphic form varies . Child gods are depicted nude , as are some adult gods when their procreative powers are emphasized . Certain male deities are given heavy bellies and breasts , signifying either androgyny or prosperity and abundance . Whereas most male gods have red skin and most goddesses are yellow — the same colors used to depict Egyptian men and women — some are given unusual , symbolic skin colors . Thus the blue skin and paunchy figure of the god Hapi alludes to the Nile flood he represents and the nourishing fertility it brought . A few deities , such as Osiris , Ptah , and Min , have a \" mummiform \" appearance , with their limbs tightly swathed in cloth . Although these gods resemble mummies , the earliest examples predate the cloth @-@ wrapped style of mummification , and this form may instead hark back to the earliest , limbless depictions of deities . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n In official writings , pharaohs are said to be divine , and they are constantly depicted in the company of the deities of the pantheon . Each pharaoh and his predecessors were considered the successors of the gods who had ruled Egypt in mythic prehistory . Living kings were equated with Horus and called the \" son \" of many deities , particularly Osiris and Ra ; deceased kings were equated with these elder gods . Pharaohs had their own mortuary temples where rituals were performed for them during their lives and after their deaths . But few pharaohs were worshipped as gods long after their lifetimes , and non @-@ official texts portray kings in a human light . For these reasons , scholars disagree about how genuinely most Egyptians believed the king to be a god . He may only have been considered divine when he was performing ceremonies . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n However much it was believed , the king 's divine status was the rationale for his role as Egypt 's representative to the gods , as he formed a link between the divine and human realms . The Egyptians believed the gods needed temples to dwell in , as well as the periodic performance of rituals and presentation of offerings to nourish them . These things were provided by the cults that the king oversaw , with their priests and laborers . Yet , according to royal ideology , temple @-@ building was exclusively the pharaoh 's work , as were the rituals that priests usually performed in his stead . These acts were a part of the king 's fundamental role : maintaining maat . The king and the nation he represented provided the gods with maat so they could continue to perform their functions , which maintained maat in the cosmos so humans could continue to live . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Although the Egyptians believed their gods to be present in the world around them , contact between the human and divine realms was mostly limited to specific circumstances . In literature , gods may appear to humans in a physical form , but in real life the Egyptians were limited to more indirect means of communication . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n The ba of a god was said to periodically leave the divine realm to dwell in the images of that god . By inhabiting these images , the gods left their concealed state and took on a physical form . To the Egyptians , a place or object that was ḏsr — \" sacred \" — was isolated and ritually pure , and thus fit for a god to inhabit . Temple statues and reliefs , as well as particular sacred animals , like the Apis bull , served as divine intermediaries in this way . Dreams and trances provided a very different venue for interaction . In these states , it was believed , people could come close to the gods and sometimes receive messages from them . Finally , according to Egyptian afterlife beliefs , human souls pass into the divine realm after death . The Egyptians therefore believed that in death they would exist on the same level as the gods and fully understand their mysterious nature . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Temples , where the state rituals were carried out , were filled with images of the gods . The most important temple image was the cult statue in the inner sanctuary . These statues were usually less than life @-@ size , and made of the same precious materials that were said to form the gods ' bodies . Many temples had several sanctuaries , each with a cult statue representing one of the gods in a group such as a family triad . The city 's primary god was envisioned as its lord , employing many of the residents as servants in the divine household that the temple represented . The gods residing in the temples of Egypt collectively represented the entire pantheon . But many deities — including some important gods as well as those that were minor or hostile — were never given temples of their own , although some were represented in the temples of other gods . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n To insulate the sacred power in the sanctuary from the impurities of the outside world , the Egyptians enclosed temple sanctuaries and greatly restricted access to them . People other than kings and high priests were thus denied contact with cult statues . The only exception was during festival processions , when the statue was carried out of the temple but still enclosed in a portable shrine . People did have less direct means of interaction . The more public parts of temples often incorporated small places for prayer , from doorways to freestanding chapels near the back of the temple building . Communities also built and managed small chapels for their own use , and some families had shrines inside their homes . Despite the gulf that separated humanity from the divine , the Egyptians were surrounded by opportunities to approach their gods . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Egyptian gods were involved in human lives as well as in the overarching order of nature . This divine influence applied mainly to Egypt , as foreign peoples were traditionally believed to be outside the divine order . But in the New Kingdom , when other nations were under Egyptian control , foreigners were said to be under the sun god 's benign rule in the same way that Egyptians were . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Thoth , as the overseer of time , was said to allot fixed lifespans to both humans and gods . Other gods were also said to govern the length of human lives , including Meskhenet , who presided over birth , and Shai , the personification of fate . Thus the time and manner of death was the main meaning of the Egyptian concept of fate , although to some extent these deities governed other events in life as well . Several texts refer to gods influencing or inspiring human decisions , working through a person 's \" heart \" — the seat of emotion and intellect in Egyptian belief . Deities were also believed to give commands , instructing the king in the governance of his realm and regulating the management of their temples . Egyptian texts rarely mention direct commands given to private persons , and these commands never evolved into a set of divinely enforced moral codes . Morality in ancient Egypt was based on the concept of maat , which , when applied to human society , meant that everyone should live in an orderly way that did not interfere with the well @-@ being of other people . Because deities were the upholders of maat , morality was connected with them . For example , the gods judged humans ' moral righteousness after death , and by the New Kingdom , a verdict of innocence in this judgment was believed to be necessary for admittance into the afterlife . But in general , morality was based on practical ways to uphold maat in daily life , rather than on strict rules that the gods laid out . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Humans had free will to ignore divine guidance and the behavior required by maat , but by doing so they could bring divine punishment upon themselves . A deity carried out this punishment using its ba , the force that manifested the god 's power in the human world . Natural disasters and human ailments were seen as the work of angry divine bas . Conversely , the gods could cure righteous people of illness or even extend their lifespans . Both these types of intervention were eventually represented by deities : Shed , who emerged in the New Kingdom to represent divine rescue from harm , and Petbe , an apotropaic god from the late eras of Egyptian history who was believed to avenge wrongdoing . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": "<|start-PI|>Humans had free will to ignore divine guidance and the behavior required by maat , but by doing so they could bring<|end-PI|> <|retrieval|> <|start-KFI|>divine punishment upon themselves<|end-KFI|> . <|start-PI|>A deity carried out this punishment using its ba ,<|end-PI|> <|retrieval|> <|start-KFI|>the force that manifested the god 's power in the human world<|end-KFI|> . <|start-PI|>Natural disasters and human ailments were seen as<|end-PI|> <|retrieval|> <|start-KFI|>the work of angry divine bas<|end-KFI|> . <|start-PI|>Conversely , the gods could cure righteous people of illness or even<|end-PI|> <|retrieval|> <|start-KFI|>extend their lifespans<|end-KFI|> . <|start-PI|>Both these types of intervention were eventually represented by deities : Shed , who emerged in the New Kingdom to represent<|end-PI|> <|retrieval|> <|start-KFI|>divine rescue from harm<|end-KFI|> , <|start-PI|>and Petbe , an apotropaic god from the late eras of Egyptian history who was believed to<|end-PI|> <|retrieval|> <|start-KFI|>avenge wrongdoing<|end-KFI|> .", "none": null, "results": [ { "Explanation": "The PI sets up the context of human free will and the potential for negative consequences, while the KFI provides the specific outcome. The <|retrieval|> token is inserted to indicate where additional information about divine punishment might be needed.", "Ideal Knowledge": "Information about the ancient Egyptian concept of maat, its importance in guiding human behavior, and the consequences of ignoring divine guidance in Egyptian religious beliefs.", "Key Factual Information": "divine punishment upon themselves", "Prefix reformulated": "What could humans bring upon themselves by ignoring divine guidance and maat?", "Prefix-intent": "Humans had free will to ignore divine guidance and the behavior required by maat , but by doing so they could bring", "Verification": "The KFI directly states the consequence of ignoring divine guidance and maat, as prompted by the PI." }, { "Explanation": "The PI introduces the concept of divine punishment through ba, and the KFI provides a definition of ba. The <|retrieval|> token is inserted to indicate where additional information about ba might be needed.", "Ideal Knowledge": "Detailed explanation of the concept of ba in ancient Egyptian religion, its role in manifesting divine power, and how it was believed to be used for punishment.", "Key Factual Information": "the force that manifested the god 's power in the human world", "Prefix reformulated": "How did a deity carry out punishment according to ancient Egyptian beliefs?", "Prefix-intent": "A deity carried out this punishment using its ba ,", "Verification": "The KFI directly explains what ba is and how it relates to divine punishment, as prompted by the PI." }, { "Explanation": "The PI sets up a query about the perception of natural disasters and ailments, while the KFI provides the specific belief. The <|retrieval|> token is inserted to indicate where additional information about divine causation might be needed.", "Ideal Knowledge": "Information on ancient Egyptian beliefs about the causes of natural disasters and illnesses, specifically their attribution to divine intervention through ba.", "Key Factual Information": "the work of angry divine bas", "Prefix reformulated": "How were natural disasters and human ailments perceived in ancient Egyptian beliefs?", "Prefix-intent": "Natural disasters and human ailments were seen as", "Verification": "The KFI directly states how natural disasters and ailments were perceived, as prompted by the PI." }, { "Explanation": "The PI introduces the concept of divine healing and sets up for additional benefits, while the KFI provides a specific benefit. The <|retrieval|> token is inserted to indicate where additional information about divine rewards might be needed.", "Ideal Knowledge": "Information on ancient Egyptian beliefs about divine intervention in human health and longevity, specifically how gods were thought to reward righteous behavior.", "Key Factual Information": "extend their lifespans", "Prefix reformulated": "What could the gods do for righteous people according to ancient Egyptian beliefs?", "Prefix-intent": "Conversely , the gods could cure righteous people of illness or even", "Verification": "The KFI directly states an additional benefit the gods could provide to righteous people, as prompted by the PI." }, { "Explanation": "The PI introduces the deity Shed and sets up a query about its representation, while the KFI provides the specific role. The <|retrieval|> token is inserted to indicate where additional information about Shed might be needed.", "Ideal Knowledge": "Detailed information about the deity Shed in ancient Egyptian religion, including its origins in the New Kingdom period and its role in protecting people from harm.", "Key Factual Information": "divine rescue from harm", "Prefix reformulated": "What did the deity Shed represent in ancient Egyptian beliefs?", "Prefix-intent": "Both these types of intervention were eventually represented by deities : Shed , who emerged in the New Kingdom to represent", "Verification": "The KFI directly states what Shed represented, as prompted by the PI." }, { "Explanation": "The PI introduces the god Petbe and sets up a query about its believed role, while the KFI provides the specific function. The <|retrieval|> token is inserted to indicate where additional information about Petbe might be needed.", "Ideal Knowledge": "Information about the god Petbe in late Egyptian religious history, including its role as an apotropaic deity and its association with avenging wrongdoing.", "Key Factual Information": "avenge wrongdoing", "Prefix reformulated": "What was the role of the god Petbe in late Egyptian history?", "Prefix-intent": "and Petbe , an apotropaic god from the late eras of Egyptian history who was believed to", "Verification": "The KFI directly states what Petbe was believed to do, as prompted by the PI." } ] }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Egyptian texts take different views on whether the gods are responsible when humans suffer unjustly . Misfortune was often seen as a product of isfet , the cosmic disorder that was the opposite of maat , and therefore the gods were not guilty of causing evil events . Some deities who were closely connected with isfet , such as Set , could be blamed for disorder within the world without placing guilt on the other gods . But some writings do accuse the deities of causing human misery , while others give theodicies in the gods ' defense . Beginning in the Middle Kingdom , several texts connected the issue of evil in the world with a myth in which the creator god fights a human rebellion against his rule and then withdraws from the earth . Because of this human misbehavior , the creator is distant from his creation , allowing suffering to exist . New Kingdom writings do not question the just nature of the gods as strongly as those of the Middle Kingdom . They emphasize humans ' direct , personal relationships with deities and the gods ' power to intervene in human events . People in this era put faith in specific gods who they hoped would help and protect them through their lives . As a result , upholding the ideals of maat grew less important than gaining the gods ' favor as a way to guarantee a good life . Even the pharaohs were regarded as dependent on divine aid , and after the New Kingdom came to an end , government was increasingly influenced by oracles communicating the gods ' will . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Official religious practices , which maintained maat for the benefit of all Egypt , were related to , but distinct from , the religious practices of ordinary people , who sought the gods ' help for their personal problems . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Official religion involved a variety of rituals , based in temples . Some rites were performed every day , whereas others were festivals , taking place at longer intervals and often limited to a particular temple or deity . The gods received their offerings in daily ceremonies , in which their statues were clothed , anointed , and presented with food as hymns were recited in their honor . These offerings , in addition to maintaining maat for the gods , celebrated deities ' life @-@ giving generosity and encouraged them to remain benevolent rather than vengeful . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": "Official religion involved a variety of rituals , based in <|start-PI|> Official religion involved a variety of rituals , based in <|end-PI|> <|retrieval|> <|start-KFI|> temples <|end-KFI|> . <|start-PI|> Some rites were performed every day , whereas others were <|end-PI|> <|retrieval|> <|start-KFI|> festivals , taking place at longer intervals and often limited to a particular temple or deity <|end-KFI|> . <|start-PI|> The gods received their offerings in daily ceremonies , in which their statues were <|end-PI|> <|retrieval|> <|start-KFI|> clothed , anointed , and presented with food as hymns were recited in their honor <|end-KFI|> . <|start-PI|> These offerings , in addition to maintaining maat for the gods , celebrated deities ' life @-@ giving generosity and <|end-PI|> <|retrieval|> <|start-KFI|> encouraged them to remain benevolent rather than vengeful <|end-KFI|> .", "none": null, "results": [ { "Explanation": "The PI introduces the concept of official religion and its rituals, while the KFI specifies where these rituals took place. The <|retrieval|> token is inserted to indicate where additional information about the religious practices might be needed.", "Ideal Knowledge": "Information about the role of temples in ancient religious practices, specifically their function as centers for various rituals in official religions.", "Key Factual Information": "temples", "Prefix reformulated": "Where were the rituals of the official religion based?", "Prefix-intent": "Official religion involved a variety of rituals , based in", "Verification": "The KFI directly provides the location of the rituals mentioned in the PI." }, { "Explanation": "The PI sets up a comparison between daily rites and other types, while the KFI provides specific information about festivals. The <|retrieval|> token is inserted to indicate where additional information about different types of religious rites might be needed.", "Ideal Knowledge": "Detailed information about the types and frequency of religious rites in ancient religions, including daily rituals and festival celebrations, along with their specific characteristics and limitations.", "Key Factual Information": "festivals , taking place at longer intervals and often limited to a particular temple or deity", "Prefix reformulated": "What were the other types of rites in the official religion besides daily rites?", "Prefix-intent": "Some rites were performed every day , whereas others were", "Verification": "The KFI directly describes the other types of rites, as prompted by the PI." }, { "Explanation": "The PI introduces the concept of daily offerings to gods, while the KFI provides specific details about the rituals performed. The <|retrieval|> token is inserted to indicate where additional information about daily religious ceremonies might be needed.", "Ideal Knowledge": "Detailed descriptions of daily religious ceremonies in ancient religions, including the specific rituals performed on statues of gods, such as clothing, anointing, and offering food, as well as the recitation of hymns.", "Key Factual Information": "clothed , anointed , and presented with food as hymns were recited in their honor", "Prefix reformulated": "What was done to the statues of gods during daily ceremonies?", "Prefix-intent": "The gods received their offerings in daily ceremonies , in which their statues were", "Verification": "The KFI directly describes the actions performed on the statues during daily ceremonies, as prompted by the PI." }, { "Explanation": "The PI sets up the known purposes of the offerings, while the KFI provides an additional purpose. The <|retrieval|> token is inserted to indicate where additional information about the purposes of religious offerings might be needed.", "Ideal Knowledge": "Information about the purposes and significance of religious offerings in ancient religions, including the concept of maat, the celebration of deities' generosity, and the belief in influencing divine behavior through offerings.", "Key Factual Information": "encouraged them to remain benevolent rather than vengeful", "Prefix reformulated": "What was the purpose of the offerings to the gods besides maintaining maat and celebrating their generosity?", "Prefix-intent": "These offerings , in addition to maintaining maat for the gods , celebrated deities ' life @-@ giving generosity and", "Verification": "The KFI directly provides an additional purpose for the offerings, as prompted by the PI." } ] }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Festivals often involved a ceremonial procession in which a cult image was carried out of the temple in a barque @-@ shaped shrine . These processions served various purposes . In Roman times , when local deities of all kinds were believed to have power over the Nile inundation , processions in many communities carried temple images to the riverbanks so the gods could invoke a large and fruitful flood . Processions also traveled between temples , as when the image of Hathor from Dendera Temple visited her consort Horus at the Temple of Edfu . Rituals for a god were often based in that deity 's mythology . Such rituals were meant to be repetitions of the events of the mythic past , renewing the beneficial effects of the original events . In the Khoiak festival in honor of Osiris , his death and resurrection were ritually reenacted at a time when crops were beginning to sprout . The returning greenery symbolized the renewal of the god 's own life . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Personal interaction with the gods took many forms . People who wanted information or advice consulted oracles , run by temples , that were supposed to convey gods ' answers to questions . Amulets and other images of protective deities were used to ward off the demons that might threaten human well @-@ being or to impart the god 's positive characteristics to the wearer . Private rituals invoked the gods ' power to accomplish personal goals , from healing sickness to cursing enemies . These practices used heka , the same force of magic that the gods used , which the creator was said to have given to humans so they could fend off misfortune . The performer of a private rite often took on the role of a god in a myth , or even threatened a deity , to involve the gods in accomplishing the goal . Such rituals coexisted with private offerings and prayers , and all three were accepted means of obtaining divine help . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Prayer and private offerings are generally called \" personal piety \" : acts that reflect a close relationship between an individual and a god . Evidence of personal piety is scant before the New Kingdom . Votive offerings and personal names , many of which are theophoric , suggest that commoners felt some connection between themselves and their gods . But firm evidence of devotion to deities became visible only in the New Kingdom , reaching a peak late in that era . Scholars disagree about the meaning of this change — whether direct interaction with the gods was a new development or an outgrowth of older traditions . Egyptians now expressed their devotion through a new variety of activities in and around temples . They recorded their prayers and their thanks for divine help on stelae . They gave offerings of figurines that represented the gods they were praying to , or that symbolized the result they desired ; thus a relief image of Hathor and a statuette of a woman could both represent a prayer for fertility . Occasionally , a person took a particular god as a patron , dedicating his or her property or labor to the god 's cult . These practices continued into the latest periods of Egyptian history . These later eras saw more religious innovations , including the practice of giving animal mummies as offerings to deities depicted in animal form , such as the cat mummies given to the feline goddess Bastet . Some of the major deities from myth and official religion were rarely invoked in popular worship , but many of the great state gods were important in popular tradition . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n The worship of some Egyptian gods spread to neighboring lands , especially to Canaan and Nubia during the New Kingdom , when those regions were under pharaonic control . In Canaan , the exported deities , including Hathor , Amun , and Set , were often syncretized with native gods , who in turn spread to Egypt . The Egyptian deities may not have had permanent temples in Canaan , and their importance there waned after Egypt lost control of the region . In contrast , many temples to the major Egyptian gods and deified pharaohs were built in Nubia . After the end of Egyptian rule there , the imported gods , particularly Amun and Isis , were syncretized with local deities and remained part of the religion of Nubia 's independent Kingdom of Kush . These gods were incorporated into the Nubian ideology of kingship much as they were in Egypt , so that Amun was considered the divine father of the king and Isis and other goddesses were linked with the Nubian queen , the kandake . Some deities reached farther . Taweret became a goddess in Minoan Crete , and Amun 's oracle at Siwa Oasis was known to and consulted by people across the Mediterranean region . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Under the Greek Ptolemaic Dynasty and then Roman rule , Greeks and Romans introduced their own deities to Egypt . These newcomers equated the Egyptian gods with their own , as part of the Greco @-@ Roman tradition of interpretatio graeca . But the worship of the native gods was not swallowed up by that of foreign ones . Instead , Greek and Roman gods were adopted as manifestations of Egyptian ones . Egyptian cults sometimes incorporated Greek language , philosophy , iconography , and even temple architecture . Meanwhile , the cults of several Egyptian deities — particularly Isis , Osiris , Anubis , the form of Horus named Harpocrates , and the fused Greco @-@ Egyptian god Serapis — were adopted into Roman religion and spread across the Roman Empire . Roman emperors , like Ptolemaic kings before them , invoked Isis and Serapis to endorse their authority , inside and outside Egypt . In the empire 's complex mix of religious traditions , Thoth was transmuted into the legendary esoteric teacher Hermes Trismegistus , and Isis , who was venerated from Britain to Mesopotamia , became the focus of a Greek @-@ style mystery cult . Isis and Hermes Trismegistus were both prominent in the Western esoteric tradition that grew from the Roman religious world . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": "<|start-PI|>Under the Greek Ptolemaic Dynasty and then Roman rule, Greeks and Romans introduced their own deities to Egypt. These newcomers equated<|end-PI|> <|retrieval|> <|start-KFI|>the Egyptian gods with their own, as part of the Greco-Roman tradition of interpretatio graeca<|end-KFI|>. <|start-PI|>But the worship of the native gods was not swallowed up by that of foreign ones. Instead,<|end-PI|> <|retrieval|> <|start-KFI|>Greek and Roman gods were adopted as manifestations of Egyptian ones<|end-KFI|>. Egyptian cults sometimes incorporated Greek language, philosophy, iconography, and even temple architecture. <|start-PI|>Meanwhile, the cults of several Egyptian deities — particularly Isis, Osiris, Anubis, the form of Horus named Harpocrates, and the fused Greco-Egyptian god Serapis — were<|end-PI|> <|retrieval|> <|start-KFI|>adopted into Roman religion and spread across the Roman Empire<|end-KFI|>. Roman emperors, like Ptolemaic kings before them, invoked Isis and Serapis to endorse their authority, inside and outside Egypt. <|start-PI|>In the empire's complex mix of religious traditions, Thoth was transmuted into<|end-PI|> <|retrieval|> <|start-KFI|>the legendary esoteric teacher Hermes Trismegistus<|end-KFI|>, and Isis, who was venerated from Britain to Mesopotamia, became the focus of a Greek-style mystery cult. <|start-PI|>Isis and Hermes Trismegistus were both prominent in<|end-PI|> <|retrieval|> <|start-KFI|>the Western esoteric tradition that grew from the Roman religious world<|end-KFI|>.", "none": null, "results": [ { "Explanation": "The PI sets up the context of Greek and Roman rule in Egypt and introduces the topic of deity introduction. The KFI provides the specific method used, namely interpretatio graeca. The <|retrieval|> token is inserted to indicate where additional information about this practice might be needed.", "Ideal Knowledge": "Information on the Greco-Roman practice of interpretatio graeca, explaining how foreign gods were equated with local deities during cultural exchanges in ancient times.", "Key Factual Information": "the Egyptian gods with their own, as part of the Greco-Roman tradition of interpretatio graeca", "Prefix reformulated": "How did Greeks and Romans introduce their deities to Egypt?", "Prefix-intent": "Under the Greek Ptolemaic Dynasty and then Roman rule, Greeks and Romans introduced their own deities to Egypt. These newcomers equated", "Verification": "The KFI directly explains how Greeks and Romans introduced their deities, as prompted by the PI." }, { "Explanation": "The PI sets up a contrast to the expected outcome of foreign deity introduction. The KFI provides the specific way in which native and foreign gods were reconciled. The <|retrieval|> token is inserted to indicate where additional information about religious syncretism might be needed.", "Ideal Knowledge": "Detailed information on the syncretism of Egyptian and Greco-Roman religions, explaining how foreign gods were incorporated into the existing Egyptian pantheon.", "Key Factual Information": "Greek and Roman gods were adopted as manifestations of Egyptian ones", "Prefix reformulated": "What happened to the worship of native Egyptian gods when foreign deities were introduced?", "Prefix-intent": "But the worship of the native gods was not swallowed up by that of foreign ones. Instead,", "Verification": "The KFI directly explains what happened to the worship of native gods, as prompted by the PI." }, { "Explanation": "The PI lists specific Egyptian deities and sets up a query about their fate. The KFI provides information on their adoption and spread within Roman religion. The <|retrieval|> token is inserted to indicate where additional information about the spread of Egyptian cults might be needed.", "Ideal Knowledge": "Information on the spread of Egyptian cults throughout the Roman Empire, detailing how these deities were incorporated into Roman religious practices and their popularity in different regions.", "Key Factual Information": "adopted into Roman religion and spread across the Roman Empire", "Prefix reformulated": "What happened to the cults of Egyptian deities like Isis, Osiris, Anubis, Harpocrates, and Serapis?", "Prefix-intent": "Meanwhile, the cults of several Egyptian deities — particularly Isis, Osiris, Anubis, the form of Horus named Harpocrates, and the fused Greco-Egyptian god Serapis — were", "Verification": "The KFI directly explains what happened to the cults of specific Egyptian deities, as prompted by the PI." }, { "Explanation": "The PI introduces the transformation of Thoth within the Roman religious context. The KFI provides the specific form into which Thoth was transmuted. The <|retrieval|> token is inserted to indicate where additional information about Hermes Trismegistus might be needed.", "Ideal Knowledge": "Detailed information on the syncretism of Thoth with Hermes, explaining the creation of Hermes Trismegistus and his significance in esoteric traditions.", "Key Factual Information": "the legendary esoteric teacher Hermes Trismegistus", "Prefix reformulated": "How was the Egyptian god Thoth transformed in the Roman Empire?", "Prefix-intent": "In the empire's complex mix of religious traditions, Thoth was transmuted into", "Verification": "The KFI directly explains how Thoth was transformed, as prompted by the PI." }, { "Explanation": "The PI introduces Isis and Hermes Trismegistus and sets up a query about their significance. The KFI provides information on the specific tradition in which they were prominent. The <|retrieval|> token is inserted to indicate where additional information about the Western esoteric tradition might be needed.", "Ideal Knowledge": "Information on the Western esoteric tradition, its origins in the Roman religious world, and the roles of Isis and Hermes Trismegistus within this tradition.", "Key Factual Information": "the Western esoteric tradition that grew from the Roman religious world", "Prefix reformulated": "In what tradition were Isis and Hermes Trismegistus prominent?", "Prefix-intent": "Isis and Hermes Trismegistus were both prominent in", "Verification": "The KFI directly explains the tradition in which Isis and Hermes Trismegistus were prominent, as prompted by the PI." } ] }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Temples and cults in Egypt itself declined as the Roman economy deteriorated in the third century AD , and beginning in the fourth century , Christians suppressed the veneration of Egyptian deities . The last formal cults , at Philae , died out in the fifth or sixth century . Most beliefs surrounding the gods themselves disappeared within a few hundred years , remaining in magical texts into the seventh and eighth centuries . But many of the practices involved in their worship , such as processions and oracles , were adapted to fit Christian ideology and persisted as part of the Coptic Church . Given the great changes and diverse influences in Egyptian culture since that time , scholars disagree about whether any modern Coptic practices are descended from those of pharaonic religion . But many festivals and other traditions of modern Egyptians , both Christian and Muslim , resemble the worship of their ancestors ' gods . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n South of Heaven is the fourth studio album by American thrash metal band Slayer . Released on July 5 , 1988 , the album was the band 's second collaboration with record producer Rick Rubin , whose production skills on Slayer 's previous album Reign in Blood had helped the band 's sound evolve . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n South of Heaven was Slayer 's second album to enter the Billboard 200 , and its last to be released by Def Jam Recordings , although the album became an American Recordings album after Rick Rubin ended his partnership with Russell Simmons . It was one of only two Def Jam titles to be distributed by Geffen Records through Warner Bros. Records because of original distributor Columbia Records ' refusal to release work by the band . The release peaked at number 57 and in 1992 was awarded a gold certification by the Recording Industry Association of America . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n In order to offset the pace of the group 's previous album , Slayer deliberately slowed down the album 's tempo . In contrast to their previous albums , the band utilized undistorted guitars and toned @-@ down vocals . While some critics praised this musical change , others — more accustomed to the style of earlier releases — were disappointed . The songs \" Mandatory Suicide \" and the title track , however , have become permanent features of the band 's live setlist . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n South of Heaven was recorded in Los Angeles , California with Reign in Blood producer Rick Rubin . PopMatters reviewer Adrien Begrand observed that Rubin 's production \" shoves [ Dave ] Lombardo 's drumming right up front in the mix . \" Guitarist Jeff Hanneman has since said that South of Heaven was the only album the band members discussed before writing the music . Aware that they \" couldn 't top Reign in Blood \" , and that whatever they recorded would be \" compared to that album \" , he believed they \" had to slow down \" , something Slayer had never done on albums before , or since . Guitarist Kerry King cited the need to \" keep people guessing \" as another reason for the musical shift . \" In order to contrast the aggressive assault put forth on Reign in Blood , Slayer consciously slowed down the tempo of the album as a whole \" , according to Slayer 's official biography . \" They also added elements like undistorted guitars and toned @-@ down vocal styles not heard on previous albums . \" \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n King has since been critical of his performance , which he describes as his \" most lackluster . \" King attributes this to the fact he had recently married , and moved to Phoenix , Arizona . Describing himself as \" probably the odd man out at that point \" , he admitted he \" didn ’ t participate as much because of that . \" Hanneman said : \" We go through dry spells sometimes , but the good thing about having two guitar players that can write music is that you are never gonna go without . I guess at that time , Kerry was hitting a dry spell . \" King has also been critical of the album in general , describing it as one of his least favorite Slayer albums . He feels vocalist Tom Araya moved too far away from his regular vocal style , and \" added too much singing . \" Drummer Dave Lombardo has since observed : \" There was fire on all the records , but it started dimming when South of Heaven came into the picture . And that 's me personally . Again , I was probably wanting something else . \" \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Judas Priest 's \" Dissident Aggressor \" is the only cover version to appear on a Slayer studio album . The song was chosen due to its war @-@ themed lyrics . Hanneman described the track as \" more just like one of those odd songs that a lot of people didn 't know , but it was a favorite of Kerry and I , so we just picked that one . \" Meanwhile , \" Cleanse the Soul \" has been heavily criticized by King who said that he hates the track : \" That 's one of the black marks in our history , in my book . I just fucking think it 's horrible . [ Laughs ] I hate the opening riff . It 's what we call a ' happy riff . ' It 's just like ' la @-@ lala @-@ la @-@ la @-@ la . ' I can 't see myself playing it , but after that , where it gets heavier , I like that section . If we ever did a medley , I 'd put part of that in there . \" The Slayer boxset Soundtrack to the Apocalypse featured , along with four songs of the album , an early version of the title track , recorded at Hanneman 's home . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": "<|start-PI|>Judas Priest's \"Dissident Aggressor\" is the only cover version to appear on a Slayer studio album<|end-PI|> <|retrieval|> <|start-KFI|>The song was chosen due to its war-themed lyrics<|end-KFI|>. <|start-PI|>Hanneman described the track as<|end-PI|> <|retrieval|> <|start-KFI|>\"more just like one of those odd songs that a lot of people didn't know, but it was a favorite of Kerry and I, so we just picked that one.\"<|end-KFI|> Meanwhile, <|start-PI|>\"Cleanse the Soul\" has been heavily criticized by King who said that<|end-PI|> <|retrieval|> <|start-KFI|>he hates the track: \"That's one of the black marks in our history, in my book. I just fucking think it's horrible. [Laughs] I hate the opening riff. It's what we call a 'happy riff.' It's just like 'la-lala-la-la-la.' I can't see myself playing it, but after that, where it gets heavier, I like that section. If we ever did a medley, I'd put part of that in there.\"<|end-KFI|> <|start-PI|>The Slayer boxset Soundtrack to the Apocalypse featured, along with four songs of the album,<|end-PI|> <|retrieval|> <|start-KFI|>an early version of the title track, recorded at Hanneman's home<|end-KFI|>.", "none": null, "results": [ { "Explanation": "The PI introduces the unique cover song on a Slayer album, and the KFI provides the reason for its selection. The <|retrieval|> token is inserted to indicate where additional information about the song choice might be needed.", "Ideal Knowledge": "Information about Slayer's album tracks, including cover versions, with details on why specific songs were chosen for inclusion.", "Key Factual Information": "The song was chosen due to its war-themed lyrics", "Prefix reformulated": "What is the only cover version to appear on a Slayer studio album?", "Prefix-intent": "Judas Priest's \"Dissident Aggressor\" is the only cover version to appear on a Slayer studio album", "Verification": "The KFI provides the reason for choosing the song, which is directly related to the PI's statement about the cover version." }, { "Explanation": "The PI sets up Hanneman's description of the track, and the KFI provides his exact quote. The <|retrieval|> token is inserted to indicate where additional context or quotes might be needed.", "Ideal Knowledge": "Quotes and interviews from Slayer band members, specifically Hanneman, discussing their song choices and album tracks.", "Key Factual Information": "\"more just like one of those odd songs that a lot of people didn't know, but it was a favorite of Kerry and I, so we just picked that one.\"", "Prefix reformulated": "How did Hanneman describe the track \"Dissident Aggressor\"?", "Prefix-intent": "Hanneman described the track as", "Verification": "The KFI directly quotes Hanneman's description of the track, as prompted by the PI." }, { "Explanation": "The PI introduces King's criticism of \"Cleanse the Soul,\" and the KFI provides his extensive quote explaining his dislike for the track. The <|retrieval|> token is inserted to indicate where additional context or quotes might be needed.", "Ideal Knowledge": "Detailed interviews or statements from Slayer band members, particularly King, discussing their opinions on specific tracks from their albums.", "Key Factual Information": "he hates the track: \"That's one of the black marks in our history, in my book. I just fucking think it's horrible. [Laughs] I hate the opening riff. It's what we call a 'happy riff.' It's just like 'la-lala-la-la-la.' I can't see myself playing it, but after that, where it gets heavier, I like that section. If we ever did a medley, I'd put part of that in there.\"", "Prefix reformulated": "What did King say about \"Cleanse the Soul\"?", "Prefix-intent": "\"Cleanse the Soul\" has been heavily criticized by King who said that", "Verification": "The KFI directly quotes King's criticism of the track, as prompted by the PI." }, { "Explanation": "The PI introduces the Slayer boxset and sets up information about its contents, while the KFI provides specific details about an additional track. The <|retrieval|> token is inserted to indicate where additional information about the boxset's contents might be needed.", "Ideal Knowledge": "Detailed track listings and special features of Slayer's boxset releases, including information about rare or early versions of songs.", "Key Factual Information": "an early version of the title track, recorded at Hanneman's home", "Prefix reformulated": "What did the Slayer boxset Soundtrack to the Apocalypse feature in addition to four songs from the album?", "Prefix-intent": "The Slayer boxset Soundtrack to the Apocalypse featured, along with four songs of the album,", "Verification": "The KFI directly provides information about an additional track featured in the boxset, as prompted by the PI." } ] }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Artist Larry Carroll and Illustrator Howard Schwartzberg designed the cover artwork for South of Heaven , having designed the artwork for Slayer 's previous album Reign in Blood . Photographer Glen E. Friedman took the promotional shot which surfaced as the back cover of South of Heaven around the time of 1986 's Reign in Blood . Lombardo felt it made Slayer seem as though they \" had matured a little bit \" , while Friedman himself deemed it \" a really cool back cover \" and \" one of the most classic shots of them [ Slayer ] ever . \" \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": "<|start-PI|>Artist Larry Carroll and Illustrator Howard Schwartzberg designed the cover artwork for<|end-PI|> <|retrieval|> <|start-KFI|>South of Heaven , having designed the artwork for Slayer 's previous album Reign in Blood<|end-KFI|>. <|start-PI|>Photographer Glen E. Friedman took the promotional shot which surfaced as<|end-PI|> <|retrieval|> <|start-KFI|>the back cover of South of Heaven around the time of 1986 's Reign in Blood<|end-KFI|>. <|start-PI|>Lombardo felt it made Slayer seem as though they<|end-PI|> <|retrieval|> <|start-KFI|>had matured a little bit<|end-KFI|>, while <|start-PI|>Friedman himself deemed it<|end-PI|> <|retrieval|> <|start-KFI|>a really cool back cover and one of the most classic shots of them [ Slayer ] ever<|end-KFI|>.", "none": null, "results": [ { "Explanation": "The PI introduces the artists and sets up a query about their work, while the KFI provides the specific album and additional context. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.", "Ideal Knowledge": "A list of Slayer albums with their cover artwork designers, specifically mentioning South of Heaven and Reign in Blood as designed by Larry Carroll and Howard Schwartzberg.", "Key Factual Information": "South of Heaven , having designed the artwork for Slayer 's previous album Reign in Blood", "Prefix reformulated": "What album did Larry Carroll and Howard Schwartzberg design the cover artwork for?", "Prefix-intent": "Artist Larry Carroll and Illustrator Howard Schwartzberg designed the cover artwork for", "Verification": "The KFI directly provides the album information requested in the PI." }, { "Explanation": "The PI introduces the photographer and his work, while the KFI provides specific details about how and when the shot was used. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.", "Ideal Knowledge": "Detailed information about Glen E. Friedman's work with Slayer, including the specific promotional shots he took and how they were used in album packaging, with dates and album titles.", "Key Factual Information": "the back cover of South of Heaven around the time of 1986 's Reign in Blood", "Prefix reformulated": "What did Glen E. Friedman's promotional shot become?", "Prefix-intent": "Photographer Glen E. Friedman took the promotional shot which surfaced as", "Verification": "The KFI directly describes where the promotional shot was used and when it was taken, as prompted by the PI." }, { "Explanation": "The PI sets up Lombardo's opinion about the promotional shot's effect, and the KFI provides his specific thoughts. The <|retrieval|> token is inserted to indicate where additional information about band members' opinions might be needed.", "Ideal Knowledge": "Quotes and opinions from Slayer band members, particularly Lombardo, about their image evolution and how different promotional materials affected public perception of the band.", "Key Factual Information": "had matured a little bit", "Prefix reformulated": "How did Lombardo feel the promotional shot made Slayer appear?", "Prefix-intent": "Lombardo felt it made Slayer seem as though they", "Verification": "The KFI directly provides Lombardo's opinion about the effect of the promotional shot, as prompted by the PI." }, { "Explanation": "The PI introduces Friedman's opinion of his own work, and the KFI provides his specific thoughts. The <|retrieval|> token is inserted to indicate where additional information about the photographer's perspective might be needed.", "Ideal Knowledge": "Quotes from Glen E. Friedman about his work with Slayer, particularly his thoughts on the promotional shot used for the South of Heaven back cover, and how it compares to his other work with the band.", "Key Factual Information": "a really cool back cover and one of the most classic shots of them [ Slayer ] ever", "Prefix reformulated": "How did Glen E. Friedman describe his promotional shot of Slayer?", "Prefix-intent": "Friedman himself deemed it", "Verification": "The KFI directly provides Friedman's opinion about his own work, as prompted by the PI." } ] }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n South of Heaven was released on July 5 , 1988 , and was the final Slayer album distributed via Def Jam Records . When label co @-@ founders Russell Simmons and Rubin parted ways , Slayer signed to Rubin 's newly founded Def American Recordings label . The album peaked at number 57 on the Billboard 200 album chart , and on November 20 , 1992 , became Slayer 's second album to be certified gold in the United States . South of Heaven was awarded silver certification in the United Kingdom on January 1 , 1993 , Slayer 's first record to do so in that country . Slayer 's official biography states that \" some critics praised the album as demonstrating Slayer 's desire to grow musically and avoid repeating themselves . \" Alex Henderson of AllMusic described the record as \" disturbing and powerful , \" while Joe Matera of Ultimate Guitar deemed the album a slight departure ; he wrote that while the pace was slowed down , it \" didn 't sacrifice any of the heaviness inherent in Slayer 's music . \" \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": "<|start-PI|>South of Heaven was released on<|end-PI|> <|retrieval|> <|start-KFI|>July 5, 1988<|end-KFI|>, and <|start-PI|>South of Heaven was the final Slayer album distributed via<|end-PI|> <|retrieval|> <|start-KFI|>Def Jam Records<|end-KFI|>. When label co-founders Russell Simmons and Rubin parted ways, Slayer signed to Rubin's newly founded Def American Recordings label. <|start-PI|>The album peaked at number<|end-PI|> <|retrieval|> <|start-KFI|>57 on the Billboard 200 album chart<|end-KFI|>, and on November 20, 1992, became Slayer's second album to be certified gold in the United States. <|start-PI|>South of Heaven was awarded silver certification in the United Kingdom on<|end-PI|> <|retrieval|> <|start-KFI|>January 1, 1993, Slayer's first record to do so in that country<|end-KFI|>. Slayer's official biography states that \"some critics praised the album as demonstrating Slayer's desire to grow musically and avoid repeating themselves.\" <|start-PI|>Alex Henderson of AllMusic described the record as<|end-PI|> <|retrieval|> <|start-KFI|>\"disturbing and powerful,\"<|end-KFI|> while <|start-PI|>Joe Matera of Ultimate Guitar deemed the album a slight departure; he wrote that while the pace was slowed down, it<|end-PI|> <|retrieval|> <|start-KFI|>\"didn't sacrifice any of the heaviness inherent in Slayer's music.\"<|end-KFI|>", "none": null, "results": [ { "Explanation": "The PI introduces the album and sets up a query about its release date, while the KFI provides the specific date. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.", "Ideal Knowledge": "A comprehensive list of Slayer's album releases, including South of Heaven, with specific release dates and record label information.", "Key Factual Information": "July 5, 1988", "Prefix reformulated": "When was Slayer's album South of Heaven released?", "Prefix-intent": "South of Heaven was released on", "Verification": "The KFI directly provides the release date requested in the PI." }, { "Explanation": "The PI sets up a query about the distribution of the album, and the KFI provides the specific record label. The <|retrieval|> token is inserted to indicate where additional information about Slayer's record label history might be needed.", "Ideal Knowledge": "A detailed history of Slayer's record label associations, focusing on their relationship with Def Jam Records and the transition to Def American Recordings.", "Key Factual Information": "Def Jam Records", "Prefix reformulated": "Which record label distributed Slayer's final album South of Heaven?", "Prefix-intent": "South of Heaven was the final Slayer album distributed via", "Verification": "The KFI directly provides the record label information requested in the PI." }, { "Explanation": "The PI sets up a query about the album's chart performance, and the KFI provides the specific peak position. The <|retrieval|> token is inserted to indicate where additional information about the album's chart history might be needed.", "Ideal Knowledge": "Billboard 200 chart history for Slayer's albums, specifically highlighting South of Heaven's performance and peak position.", "Key Factual Information": "57 on the Billboard 200 album chart", "Prefix reformulated": "What was the peak position of Slayer's South of Heaven album on the Billboard 200 chart?", "Prefix-intent": "The album peaked at number", "Verification": "The KFI directly provides the peak chart position requested in the PI." }, { "Explanation": "The PI sets up a query about the album's certification in the UK, and the KFI provides the specific date and its significance. The <|retrieval|> token is inserted to indicate where additional information about the album's certifications might be needed.", "Ideal Knowledge": "A comprehensive list of Slayer's album certifications in various countries, with specific dates and significance of each certification.", "Key Factual Information": "January 1, 1993, Slayer's first record to do so in that country", "Prefix reformulated": "When did Slayer's South of Heaven album receive silver certification in the UK?", "Prefix-intent": "South of Heaven was awarded silver certification in the United Kingdom on", "Verification": "The KFI directly provides the certification date and additional context requested in the PI." }, { "Explanation": "The PI introduces a critic's opinion, and the KFI provides the specific description. The <|retrieval|> token is inserted to indicate where additional information about critical reception might be needed.", "Ideal Knowledge": "A collection of critical reviews for Slayer's South of Heaven album, focusing on Alex Henderson's AllMusic review and other notable critic opinions.", "Key Factual Information": "disturbing and powerful", "Prefix reformulated": "How did Alex Henderson of AllMusic describe Slayer's South of Heaven album?", "Prefix-intent": "Alex Henderson of AllMusic described the record as", "Verification": "The KFI directly provides the description given by Alex Henderson, as requested in the PI." }, { "Explanation": "The PI introduces Joe Matera's opinion on the album's departure from previous works, and the KFI provides the specific description of how the album maintained its heaviness despite the change. The <|retrieval|> token is inserted to indicate where additional information about the album's musical style might be needed.", "Ideal Knowledge": "A detailed analysis of Slayer's musical evolution, focusing on the changes in tempo and style in South of Heaven compared to their previous works, with quotes from music critics and reviewers.", "Key Factual Information": "didn't sacrifice any of the heaviness inherent in Slayer's music", "Prefix reformulated": "How did Joe Matera of Ultimate Guitar describe the change in Slayer's South of Heaven album?", "Prefix-intent": "Joe Matera of Ultimate Guitar deemed the album a slight departure; he wrote that while the pace was slowed down, it", "Verification": "The KFI directly provides Joe Matera's description of the album's change, as prompted by the PI." } ] }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Reviewing the 2003 Slayer box set Soundtrack to the Apocalypse , Adrien Begrand of PopMatters described the album as \" their most underrated , and on this set , its five selections show how highly the band thinks of the record . \" KNAC.com 's Peter Atkinson was also positive , saying the album has a \" grandiosity and imposing presence \" which makes the record \" so magnificent . \" Grave 's Ola Lindgren and Bolt Thrower 's Karl Willetts both rate South of Heaven as amongst the top five albums of all time , while Max Kolesne of Brazilian death metal group Krisiun remembers hearing the song \" Silent Scream \" for the first time : \" It just blew me away . It was like fast double @-@ bass , fast kicks during the whole song . That was very inspiring for me . \" When discussing Slayer in an October 2007 interview , Evile frontman Matt Drake stated that while Reign in Blood \" was just speed \" , South of Heaven proved that the group could write \" slow material as well . \" Metal Forces reviewer gives \" the band credit for at least making an effort to try something new and not being afraid to experiment at such a crucial stage of their career \" , creating \" one of the more original sounding thrash / speed metal albums he heard in a long while \" . He remarks , however , that \" if you ’ re expecting to hear Reign in Blood Part Two , you ’ ll be in for a major disappointment \" . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": "Reviewing the 2003 Slayer box set Soundtrack to the Apocalypse , <|start-PI|>Adrien Begrand of PopMatters described the album as<|end-PI|> <|retrieval|> <|start-KFI|>their most underrated , and on this set , its five selections show how highly the band thinks of the record<|end-KFI|> . <|start-PI|>KNAC.com 's Peter Atkinson was also positive , saying the album has<|end-PI|> <|retrieval|> <|start-KFI|>a grandiosity and imposing presence which makes the record so magnificent<|end-KFI|> . <|start-PI|>Grave 's Ola Lindgren and Bolt Thrower 's Karl Willetts both rate South of Heaven as<|end-PI|> <|retrieval|> <|start-KFI|>amongst the top five albums of all time<|end-KFI|> , while <|start-PI|>Max Kolesne of Brazilian death metal group Krisiun remembers hearing the song Silent Scream for the first time<|end-PI|> <|retrieval|> <|start-KFI|>It just blew me away . It was like fast double @-@ bass , fast kicks during the whole song . That was very inspiring for me<|end-KFI|> . <|start-PI|>When discussing Slayer in an October 2007 interview , Evile frontman Matt Drake stated that while Reign in Blood was just speed , South of Heaven proved that the group could<|end-PI|> <|retrieval|> <|start-KFI|>write slow material as well<|end-KFI|> . <|start-PI|>Metal Forces reviewer gives the band credit for<|end-PI|> <|retrieval|> <|start-KFI|>at least making an effort to try something new and not being afraid to experiment at such a crucial stage of their career , creating one of the more original sounding thrash / speed metal albums he heard in a long while<|end-KFI|> . He remarks , however , that if you ' re expecting to hear Reign in Blood Part Two , you ' ll be in for a major disappointment .", "none": null, "results": [ { "Explanation": "The PI introduces the review and the reviewer, setting up for the specific description, while the KFI provides the actual description. The <|retrieval|> token is inserted to indicate where additional information about the review might be needed.", "Ideal Knowledge": "A comprehensive review of the 2003 Slayer box set Soundtrack to the Apocalypse, focusing on Adrien Begrand's analysis and the significance of the album's inclusion in the set.", "Key Factual Information": "their most underrated , and on this set , its five selections show how highly the band thinks of the record", "Prefix reformulated": "How did Adrien Begrand of PopMatters describe the album in the 2003 Slayer box set Soundtrack to the Apocalypse?", "Prefix-intent": "Reviewing the 2003 Slayer box set Soundtrack to the Apocalypse , Adrien Begrand of PopMatters described the album as", "Verification": "The KFI directly provides the description of the album as stated by Adrien Begrand, which is prompted by the PI." }, { "Explanation": "The PI introduces another reviewer and indicates a positive review, while the KFI provides the specific qualities mentioned. The <|retrieval|> token is inserted to indicate where additional information about this review might be needed.", "Ideal Knowledge": "A detailed review by Peter Atkinson from KNAC.com, focusing on the album's qualities and overall impact, particularly its grandiosity and presence.", "Key Factual Information": "a grandiosity and imposing presence which makes the record so magnificent", "Prefix reformulated": "What did Peter Atkinson from KNAC.com say about the album's qualities?", "Prefix-intent": "KNAC.com 's Peter Atkinson was also positive , saying the album has", "Verification": "The KFI directly provides Peter Atkinson's positive description of the album, as prompted by the PI." }, { "Explanation": "The PI introduces two musicians and sets up for their opinion on the album, while the KFI provides their specific rating. The <|retrieval|> token is inserted to indicate where additional information about these musicians' opinions might be needed.", "Ideal Knowledge": "Opinions and rankings of influential metal musicians regarding Slayer's albums, specifically focusing on South of Heaven and its place among the greatest albums of all time.", "Key Factual Information": "amongst the top five albums of all time", "Prefix reformulated": "How do Ola Lindgren from Grave and Karl Willetts from Bolt Thrower rate the album South of Heaven?", "Prefix-intent": "Grave 's Ola Lindgren and Bolt Thrower 's Karl Willetts both rate South of Heaven as", "Verification": "The KFI directly provides the rating given by the two musicians, as prompted by the PI." }, { "Explanation": "The PI introduces Max Kolesne and sets up for his reaction to the song, while the KFI provides his specific description and emotional response. The <|retrieval|> token is inserted to indicate where additional information about this musician's experience might be needed.", "Ideal Knowledge": "Detailed accounts from metal musicians about their first experiences with influential songs, focusing on the technical aspects and emotional impact of Slayer's Silent Scream.", "Key Factual Information": "It just blew me away . It was like fast double @-@ bass , fast kicks during the whole song . That was very inspiring for me", "Prefix reformulated": "What was Max Kolesne's reaction to hearing the song Silent Scream for the first time?", "Prefix-intent": "Max Kolesne of Brazilian death metal group Krisiun remembers hearing the song Silent Scream for the first time", "Verification": "The KFI directly provides Max Kolesne's reaction and description of the song, as prompted by the PI." }, { "Explanation": "The PI introduces Matt Drake's comparison of two Slayer albums, while the KFI provides the specific contrast he noted. The <|retrieval|> token is inserted to indicate where additional information about this comparison might be needed.", "Ideal Knowledge": "A comprehensive comparison of Slayer's albums Reign in Blood and South of Heaven, focusing on the evolution of their musical style and songwriting capabilities.", "Key Factual Information": "write slow material as well", "Prefix reformulated": "What did Evile frontman Matt Drake say about Slayer's South of Heaven compared to Reign in Blood?", "Prefix-intent": "When discussing Slayer in an October 2007 interview , Evile frontman Matt Drake stated that while Reign in Blood was just speed , South of Heaven proved that the group could", "Verification": "The KFI directly provides Matt Drake's comparison between the two albums, as prompted by the PI." }, { "Explanation": "The PI introduces the Metal Forces review and sets up for their positive comments, while the KFI provides the specific praise and reasoning. The <|retrieval|> token is inserted to indicate where additional information about this review might be needed.", "Ideal Knowledge": "A detailed review from Metal Forces magazine, focusing on Slayer's willingness to experiment and innovate within the thrash/speed metal genre, particularly with South of Heaven.", "Key Factual Information": "at least making an effort to try something new and not being afraid to experiment at such a crucial stage of their career , creating one of the more original sounding thrash / speed metal albums he heard in a long while", "Prefix reformulated": "What did the Metal Forces reviewer give Slayer credit for regarding South of Heaven?", "Prefix-intent": "Metal Forces reviewer gives the band credit for", "Verification": "The KFI directly provides the reviewer's praise for the band's efforts and the resulting album, as prompted by the PI." } ] }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Kim Neely of Rolling Stone dismissed the album as \" genuinely offensive satanic drivel . \" Slayer 's official biography states : \" The new sounds disappointed some of the band 's fans who were more accustomed to the style of earlier releases . \" Michael Roberts of Westworld Online said this was due to some of the numbers moving \" at the sludgier speed of Black Sabbath . \" Araya commented that the \" album was a late bloomer — it wasn 't really received well , but it kind of grew on everybody later . \" \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n The title track and the song \" Mandatory Suicide \" have received various cover interpretations , particularly on Slayer tribute albums . Toni Ferguson recorded string quartet adaptations of both tracks on the album The String Quartet Tribute to Slayer : The Evil You Dread , with the former cover being described as having \" menacing chord shifts \" by AllMusic 's Johnny Loftus . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": "<|start-PI|>The title track and the song \" Mandatory Suicide \" have received various cover interpretations , particularly on<|end-PI|> <|retrieval|> <|start-KFI|>Slayer tribute albums<|end-KFI|>. <|start-PI|>Toni Ferguson recorded string quartet adaptations of both tracks on the album<|end-PI|> <|retrieval|> <|start-KFI|>The String Quartet Tribute to Slayer : The Evil You Dread<|end-KFI|> , <|start-PI|>with the former cover being described as having<|end-PI|> <|retrieval|> <|start-KFI|>\" menacing chord shifts \" by AllMusic 's Johnny Loftus<|end-KFI|>.", "none": null, "results": [ { "Explanation": "The PI introduces the songs and sets up a query about their cover interpretations, while the KFI specifies where these covers can be found. The <|retrieval|> token is inserted to indicate where additional information about the tribute albums might be needed.", "Ideal Knowledge": "A list of Slayer tribute albums featuring cover versions of the title track and \"Mandatory Suicide\", including details about the artists performing the covers and the release dates of these albums.", "Key Factual Information": "Slayer tribute albums", "Prefix reformulated": "Where have the title track and the song \"Mandatory Suicide\" received various cover interpretations?", "Prefix-intent": "The title track and the song \" Mandatory Suicide \" have received various cover interpretations , particularly on", "Verification": "The KFI directly provides information about where the cover interpretations can be found, as prompted by the PI." }, { "Explanation": "The PI introduces Toni Ferguson's recordings and sets up a query about the album, while the KFI provides the specific album title. The <|retrieval|> token is inserted to indicate where additional information about the album might be needed.", "Ideal Knowledge": "Detailed information about the album \"The String Quartet Tribute to Slayer: The Evil You Dread\", including its track list, release date, and any critical reception or reviews it received.", "Key Factual Information": "The String Quartet Tribute to Slayer : The Evil You Dread", "Prefix reformulated": "What album did Toni Ferguson record string quartet adaptations of both tracks on?", "Prefix-intent": "Toni Ferguson recorded string quartet adaptations of both tracks on the album", "Verification": "The KFI directly provides the name of the album where Toni Ferguson's string quartet adaptations can be found, as prompted by the PI." }, { "Explanation": "The PI sets up a query about the description of the cover, and the KFI provides the specific musical characteristic and its source. The <|retrieval|> token is inserted to indicate where additional information about the musical analysis might be needed.", "Ideal Knowledge": "A detailed review or analysis of Toni Ferguson's string quartet adaptation of the title track, focusing on its musical characteristics, particularly the chord progressions and their effect on the overall sound.", "Key Factual Information": "\" menacing chord shifts \" by AllMusic 's Johnny Loftus", "Prefix reformulated": "How is the former cover described in terms of its musical characteristics?", "Prefix-intent": "with the former cover being described as having", "Verification": "The KFI directly provides a description of the musical characteristics of the cover, as prompted by the PI." } ] }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n 1995 Slayer tribute album Slatanic Slaughter featured three tracks which originally appeared on South of Heaven , with the title track , \" Mandatory Suicide \" and \" Spill the Blood \" interpreted by Cemetary , Crown of Thorns and Grope respectively . Its 1998 follow up Slatanic Slaughter , Vol . 2 only featured two tracks originally from the album ; namely \" Silent Scream \" arranged by Vader and \" Read Between the Lies \" interpreted by Anathema . 1999 's Straight to Hell : A Tribute to Slayer collected four Slayer renditions which originated on the album , with versions of South of Heaven performed by Abaddon ( Venom ) and Electric Hellfire Club , \" Mandatory Suicide \" cut by Chapter 7 and \" Behind the Crooked Cross \" adapted by Gigantor . 2006 Argentine tribute album Al Sur Del Abismo ( Tributo Argentino A Slayer ) saw Nafak and Climatic Terra also respectively cover \" South of Heaven \" and \" Mandatory Suicide \" . Hatebreed covered the song \" Ghosts of War \" for their 2009 cover album For the Lions . They released a music video for it also . Korn has covered the title track at least twice live , once with Kid Rock on vocals and another using the intro to follow into one of their songs live . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n The title track itself has also been covered by Integrity 2000 , Modest Mouse and Califone , Pro @-@ Pain , and Universe Eye . Polish death metal band Decapitated covered the song \" Mandatory Suicide \" on their first full @-@ length album Winds of Creation . In 2003 , \" Silent Scream \" was covered by Children of Bodom for their album Hate Crew Deathroll in his UK version . Hardcore Punk band , The Mongoloids opened their set with the beginning of \" South of Heaven \" at Heartfest 7 on May 4 , 2013 \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": "<|start-PI|>The title track itself has also been covered by<|end-PI|> <|retrieval|> <|start-KFI|>Integrity 2000 , Modest Mouse and Califone , Pro @-@ Pain , and Universe Eye<|end-KFI|> . <|start-PI|>Polish death metal band Decapitated covered the song<|end-PI|> <|retrieval|> <|start-KFI|>\" Mandatory Suicide \" on their first full @-@ length album Winds of Creation<|end-KFI|> . <|start-PI|>In 2003 , \" Silent Scream \" was covered by Children of Bodom for their album<|end-PI|> <|retrieval|> <|start-KFI|>Hate Crew Deathroll in his UK version<|end-KFI|> . <|start-PI|>Hardcore Punk band , The Mongoloids opened their set with the beginning of<|end-PI|> <|retrieval|> <|start-KFI|>\" South of Heaven \" at Heartfest 7 on May 4 , 2013<|end-KFI|>", "none": null, "results": [ { "Explanation": "The PI introduces the topic of cover versions of the title track, and the KFI provides a list of artists who have performed covers. The <|retrieval|> token is inserted to indicate where additional information about these cover versions might be needed.", "Ideal Knowledge": "A comprehensive list of artists who have covered the title track of the album in question, including details about their versions and release dates.", "Key Factual Information": "Integrity 2000 , Modest Mouse and Califone , Pro @-@ Pain , and Universe Eye", "Prefix reformulated": "Which artists have covered the title track of the album?", "Prefix-intent": "The title track itself has also been covered by", "Verification": "The KFI directly lists the artists who have covered the title track, as prompted by the PI." }, { "Explanation": "The PI introduces Decapitated and sets up a query about their cover song, while the KFI provides the specific song title and album information. The <|retrieval|> token is inserted to indicate where additional information about this cover might be needed.", "Ideal Knowledge": "Detailed information about Decapitated's cover of 'Mandatory Suicide', including the original artist, the album it appeared on, and any notable differences in their version.", "Key Factual Information": "\" Mandatory Suicide \" on their first full @-@ length album Winds of Creation", "Prefix reformulated": "Which song did the Polish death metal band Decapitated cover?", "Prefix-intent": "Polish death metal band Decapitated covered the song", "Verification": "The KFI directly provides the song title and album information for Decapitated's cover, as prompted by the PI." }, { "Explanation": "The PI introduces Children of Bodom's cover of 'Silent Scream' and sets up a query about the album, while the KFI provides the specific album name and version. The <|retrieval|> token is inserted to indicate where additional information about this cover and album might be needed.", "Ideal Knowledge": "Information about Children of Bodom's 2003 album 'Hate Crew Deathroll', specifically focusing on their cover of 'Silent Scream' and any differences between the UK version and other releases.", "Key Factual Information": "Hate Crew Deathroll in his UK version", "Prefix reformulated": "On which album did Children of Bodom cover 'Silent Scream' in 2003?", "Prefix-intent": "In 2003 , \" Silent Scream \" was covered by Children of Bodom for their album", "Verification": "The KFI directly provides the album name for Children of Bodom's cover of 'Silent Scream', as prompted by the PI." }, { "Explanation": "The PI introduces The Mongoloids and sets up a query about their set opener, while the KFI provides the specific song title, event name, and date. The <|retrieval|> token is inserted to indicate where additional information about this performance might be needed.", "Ideal Knowledge": "Detailed information about Heartfest 7, including the full lineup, The Mongoloids' setlist, and any significance of their choice to open with 'South of Heaven'.", "Key Factual Information": "\" South of Heaven \" at Heartfest 7 on May 4 , 2013", "Prefix reformulated": "Which song did The Mongoloids open their set with at Heartfest 7?", "Prefix-intent": "Hardcore Punk band , The Mongoloids opened their set with the beginning of", "Verification": "The KFI directly provides the song title, event name, and date for The Mongoloids' performance, as prompted by the PI." } ] }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Two songs taken from the album ( \" Mandatory Suicide \" and \" South of Heaven \" ) have become near constant fixtures in the band 's live setlist , notching up appearances on the following : the live DVDs Live Intrusion , War at the Warfield , Still Reigning , Soundtrack to the Apocalypse 's deluxe edition 's bonus live disc , and the live double album Decade of Aggression . Lombardo guested with Finnish cellist group Apocalyptica on a live medley of the two tracks at 1998 's Headbanger 's Heaven festival in the Netherlands . Adrien Begrand of PopMatters described \" South of Heaven \" as \" an unorthodox set opener in theory \" , noting \" the song went over like a megaton bomb detonating the place : dozens of inverted crosses projected behind the high drum riser , the sinewy opening notes kicked in , followed by an overture of bass , cymbal crashes , and tom fills , leading up to the slowly building crescendo \" in a concert review . Lombardo remembers listening to a live rendition of \" South of Heaven \" and thinking \" ‘ Man ! There 's just so much groove in that song . ’ To my kids I was saying , ‘ Listen to that ! Listen to how groovy that is ! ’ And it 's heavy . \" A rare live version of the track featured on the JÄGERMUSIC Rarities 2004 promotional CD , given away to attendees at the Spring 2004 Jägermeister Music Tour . A live rendition of \" South of Heaven \" was also included on a bonus DVD which came with the group 's 2007 re @-@ release of ninth studio album Christ Illusion , shot in Vancouver , British Columbia during 2006 's Unholy Alliance tour . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n \" Behind the Crooked Cross \" is rarely played live as Hanneman hates the track , though King has always wanted to play it \" because it 's got a cool intro \" despite it not being his favorite song . King said \" that 's fine \" when speaking of the situation , noting \" there are songs that he wants to play that I always shoot down . \" \" Ghosts of War \" isn 't King 's favorite song either , which he attests \" everybody always wants to hear \" performed live . He confessed ; \" I like the ending , you know , I like the big heavy part and I always say , ‘ Let 's put the heavy ending at the end of \" Chemical Warfare \" and just do the last half . ’ But I could never make that fly . \" \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Slayer has toyed with the idea of creating a live set mixed with selections from the album and 1990 's Seasons in the Abyss , though Hanneman said it 's something which hasn 't been \" seriously considered . \" Metal Maniacs asked Slayer in a 2006 interview whether they would consider playing South of Heaven in the footsteps of the Still Reigning tour , to which Araya replied , \" It 's becoming a trendy thing now . I don 't know . We have some really cool albums , but I don 't think we 'll ever do that again . \" King was equally unsure , commenting , \" Probably not . And I just don 't like enough songs off South of Heaven . \" \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": "Slayer has toyed with the idea of creating a live set mixed with selections from the album and <|start-PI|> Slayer has toyed with the idea of creating a live set mixed with selections from the album and <|end-PI|> <|retrieval|> <|start-KFI|> 1990's Seasons in the Abyss <|end-KFI|>, though Hanneman said it's something which hasn't been \"seriously considered.\" <|start-PI|> Metal Maniacs asked Slayer in a 2006 interview whether they would consider playing South of Heaven in the footsteps of the Still Reigning tour, to which Araya replied, <|end-PI|> <|retrieval|> <|start-KFI|> \"It's becoming a trendy thing now. I don't know. We have some really cool albums, but I don't think we'll ever do that again.\" <|end-KFI|> <|start-PI|> King was equally unsure, commenting, <|end-PI|> <|retrieval|> <|start-KFI|> \"Probably not. And I just don't like enough songs off South of Heaven.\" <|end-KFI|>", "none": null, "results": [ { "Explanation": "The PI introduces Slayer's consideration for a live set, while the KFI provides the specific album they're considering. The <|retrieval|> token is inserted to indicate where additional information about the album might be needed.", "Ideal Knowledge": "A discography of Slayer, highlighting their 1990 album 'Seasons in the Abyss' and its significance in their live performances or potential set lists.", "Key Factual Information": "1990's Seasons in the Abyss", "Prefix reformulated": "What album has Slayer considered mixing with selections for a live set?", "Prefix-intent": "Slayer has toyed with the idea of creating a live set mixed with selections from the album and", "Verification": "The KFI directly provides the album information that Slayer has considered mixing for a live set, as prompted by the PI." }, { "Explanation": "The PI sets up the context of the interview question, while the KFI provides Araya's direct response. The <|retrieval|> token is inserted to indicate where additional information about Slayer's interview responses might be needed.", "Ideal Knowledge": "A compilation of Slayer's interview responses regarding their album performances, specifically focusing on their views about playing entire albums live and their thoughts on the 'South of Heaven' album.", "Key Factual Information": "\"It's becoming a trendy thing now. I don't know. We have some really cool albums, but I don't think we'll ever do that again.\"", "Prefix reformulated": "What was Araya's response when asked if Slayer would consider playing South of Heaven like the Still Reigning tour?", "Prefix-intent": "Metal Maniacs asked Slayer in a 2006 interview whether they would consider playing South of Heaven in the footsteps of the Still Reigning tour, to which Araya replied,", "Verification": "The KFI directly provides Araya's response to the question about playing South of Heaven, as prompted by the PI." }, { "Explanation": "The PI introduces King's response, while the KFI provides his direct comment. The <|retrieval|> token is inserted to indicate where additional information about King's opinions on Slayer's albums might be needed.", "Ideal Knowledge": "A collection of Slayer band members' opinions on their albums, particularly focusing on Kerry King's views about 'South of Heaven' and his preferences for live performances.", "Key Factual Information": "\"Probably not. And I just don't like enough songs off South of Heaven.\"", "Prefix reformulated": "What was King's comment about playing South of Heaven live?", "Prefix-intent": "King was equally unsure, commenting,", "Verification": "The KFI directly provides King's comment about playing South of Heaven live, as prompted by the PI." } ] }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n General aviation in the United Kingdom has been defined as a civil aircraft operation other than a commercial air transport flight operating to a schedule or military aviation . Although the International Civil Aviation Organization ( ICAO ) excludes any form of remunerated aviation from its definition , some commercial operations are often included within the scope of general aviation ( GA ) in the UK . The sector operates business jets , rotorcraft , piston and jet @-@ engined fixed @-@ wing aircraft , gliders of all descriptions , and lighter than air craft . Public transport operations include business ( or corporate ) aviation and air taxi services , and account for nearly half of the economic contribution made by the sector . Other commercial GA activities are aerial work , such as surveying and air ambulances , and flight training , which plays an important role in the supply of pilots to the commercial air transport ( CAT ) industry . Private flying is conducted for personal transport and recreation . It includes a strong vintage aircraft movement , and encompasses a range of air sports , such as racing , aerobatics , and parachuting , at which British teams and individuals have succeeded in international competition . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": "<|start-PI|>General aviation in the United Kingdom has been defined as<|end-PI|> <|retrieval|> <|start-KFI|>a civil aircraft operation other than a commercial air transport flight operating to a schedule or military aviation<|end-KFI|>. Although the International Civil Aviation Organization ( ICAO ) excludes any form of remunerated aviation from its definition , some commercial operations are often included within the scope of general aviation ( GA ) in the UK . <|start-PI|>The sector operates<|end-PI|> <|retrieval|> <|start-KFI|>business jets , rotorcraft , piston and jet @-@ engined fixed @-@ wing aircraft , gliders of all descriptions , and lighter than air craft<|end-KFI|>. <|start-PI|>Public transport operations include<|end-PI|> <|retrieval|> <|start-KFI|>business ( or corporate ) aviation and air taxi services , and account for nearly half of the economic contribution made by the sector<|end-KFI|>. <|start-PI|>Other commercial GA activities are<|end-PI|> <|retrieval|> <|start-KFI|>aerial work , such as surveying and air ambulances , and flight training , which plays an important role in the supply of pilots to the commercial air transport ( CAT ) industry<|end-KFI|>. <|start-PI|>Private flying is conducted for<|end-PI|> <|retrieval|> <|start-KFI|>personal transport and recreation . It includes a strong vintage aircraft movement , and encompasses a range of air sports , such as racing , aerobatics , and parachuting , at which British teams and individuals have succeeded in international competition<|end-KFI|>.", "none": null, "results": [ { "Explanation": "The PI introduces the topic of general aviation in the UK and sets up a query about its definition, while the KFI provides the specific definition. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.", "Ideal Knowledge": "A comprehensive definition of general aviation in the United Kingdom, including its scope and distinctions from commercial and military aviation.", "Key Factual Information": "a civil aircraft operation other than a commercial air transport flight operating to a schedule or military aviation", "Prefix reformulated": "What is the definition of general aviation in the United Kingdom?", "Prefix-intent": "General aviation in the United Kingdom has been defined as", "Verification": "The KFI directly provides the definition of general aviation in the UK as requested in the PI." }, { "Explanation": "The PI sets up a query about the types of aircraft operated in the sector, and the KFI provides a comprehensive list. The <|retrieval|> token is inserted to indicate where additional information about these aircraft types might be needed.", "Ideal Knowledge": "A detailed list of aircraft types used in general aviation in the UK, including specifications and common uses for each type.", "Key Factual Information": "business jets, rotorcraft, piston and jet-engined fixed-wing aircraft, gliders of all descriptions, and lighter than air craft", "Prefix reformulated": "What types of aircraft does the general aviation sector in the UK operate?", "Prefix-intent": "The sector operates", "Verification": "The KFI directly lists the types of aircraft operated in the general aviation sector, as prompted by the PI." }, { "Explanation": "The PI introduces public transport operations and sets up a query about what they include, while the KFI provides specific examples and their economic importance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support these statements.", "Ideal Knowledge": "Detailed information on public transport operations within UK general aviation, including types of services, economic impact, and market share.", "Key Factual Information": "business (or corporate) aviation and air taxi services, and account for nearly half of the economic contribution made by the sector", "Prefix reformulated": "What does public transport operations in UK general aviation include?", "Prefix-intent": "Public transport operations include", "Verification": "The KFI directly answers what public transport operations include and their economic significance, as prompted by the PI." }, { "Explanation": "The PI sets up a query about other commercial activities in GA, and the KFI provides specific examples and their importance. The <|retrieval|> token is inserted to indicate where additional information about these activities might be needed.", "Ideal Knowledge": "A comprehensive list of commercial activities within UK general aviation, including detailed descriptions of each activity and their roles in the broader aviation industry.", "Key Factual Information": "aerial work, such as surveying and air ambulances, and flight training, which plays an important role in the supply of pilots to the commercial air transport (CAT) industry", "Prefix reformulated": "What are other commercial activities in UK general aviation besides public transport?", "Prefix-intent": "Other commercial GA activities are", "Verification": "The KFI directly lists other commercial GA activities and their significance, as prompted by the PI." }, { "Explanation": "The PI sets up a query about the purposes of private flying, and the KFI provides specific purposes and examples of activities. The <|retrieval|> token is inserted to indicate where additional information about these activities and achievements might be needed.", "Ideal Knowledge": "Detailed information on private flying activities in UK general aviation, including popular recreational activities, vintage aircraft preservation, and British achievements in international air sports competitions.", "Key Factual Information": "personal transport and recreation. It includes a strong vintage aircraft movement, and encompasses a range of air sports, such as racing, aerobatics, and parachuting, at which British teams and individuals have succeeded in international competition", "Prefix reformulated": "What are the purposes of private flying in UK general aviation?", "Prefix-intent": "Private flying is conducted for", "Verification": "The KFI directly answers the purposes of private flying and provides examples of activities, as prompted by the PI." } ] }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Of the 21 @,@ 000 civil aircraft registered in the UK , 96 per cent are engaged in GA operations , and annually the GA fleet accounts for between 1 @.@ 25 and 1 @.@ 35 million hours flown . The single most common class of aircraft is the fixed @-@ wing light aircraft associated with traditional GA , but the main area of growth over the last 20 years has been in the use of more affordable aircraft , such as microlights , amateur built aeroplanes , and smaller helicopters . There are 28 @,@ 000 Private Pilot Licence holders , and 10 @,@ 000 certified glider pilots . Some of the 19 @,@ 000 pilots who hold professional licences are also engaged in GA activities . Although GA operates from more than 1 @,@ 800 aerodromes and landing sites , ranging in size from large regional airports to farm strips , over 80 per cent of GA activity is conducted at 134 of the larger aerodromes . The GA industry , which is around 7 per cent the size of its CAT cousin , employs 12 @,@ 000 people , and contributes £ 1 @.@ 4 billion to the UK economy . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n GA is regulated by the Civil Aviation Authority ( CAA ) , although regulatory powers are being increasingly transferred to the European Aviation Safety Agency ( EASA ) . The main focus is on standards of airworthiness and pilot licensing , and the objective is to promote high standards of safety . At the lighter end of the GA spectrum some regulatory authority is devolved to representative bodies , and gliding is in transition from a self @-@ regulatory model to more formal governance by EASA . Airspace regulation necessary to protect an increasing number of CAT operations has reduced the area in which GA flights can be freely conducted . The growth in CAT is also making access to larger airports more difficult for the GA sector , and smaller aerodromes are vulnerable to closure and re @-@ development for more profitable uses . The UK planning system has no remit to consider the national significance of GA public transport operations , and generally does not favour the development of smaller aerodromes catering to the GA market . The planning process has become a mechanism for addressing local aerodrome @-@ related environmental issues which , particularly regarding noise , are the main subjects of public criticism levelled at GA . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n The International Civil Aviation Organization ( ICAO ) defines general aviation ( GA ) as \" an aircraft operation other than a commercial air transport operation or an aerial work operation . \" It defines commercial air transport ( CAT ) as \" an aircraft operation involving the transport of passengers , cargo or mail for remuneration or hire \" , and aerial work as \" an aircraft operation in which an aircraft is used for specialized services such as agriculture , construction , photography , surveying , observation and patrol , search and rescue , aerial advertisement , etc . \" \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Organisations in the United Kingdom ( UK ) describe GA in less restrictive terms that include elements of commercial aviation . The British Business and General Aviation Association interprets it to be \" all aeroplane and helicopter flying except that performed by the major airlines and the Armed Services \" . The General Aviation Awareness Council applies the description \" all Civil Aviation operations other than scheduled air services and non @-@ scheduled air transport operations for remuneration or hire \" . For the purposes of a strategic review of GA in the UK , the Civil Aviation Authority ( CAA ) defined the scope of GA as \" a civil aircraft operation other than a commercial air transport flight operating to a schedule \" , and considered it necessary to depart from the ICAO definition and include aerial work and minor CAT operations . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n The first aerodrome in the UK was established by the Aero Club at Muswell Manor on the Isle of Sheppey , and in May 1909 it was the venue of the first flight conducted in the country by a British pilot , John Moore @-@ Brabazon . In 1910 the Aero Club was granted the Royal prefix , took responsibility for controlling all private flying in the UK , and started issuing the first British pilot licences . The introduction of the de Havilland DH.60 Moth in 1925 revolutionised light aviation , and the Royal Aero Club , recognising the \" vital necessity of promoting civil flying \" , formed the Light Aeroplane Club scheme . Between 1925 and 1939 around 60 flying clubs were started , and more than 5 @,@ 000 pilots were trained . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n During World War II civil aerodromes were taken over for military use , existing military airfields were expanded , and new ones were built . This resulted in a significant inventory of facilities becoming available after the war . Pre @-@ war civil aerodromes , for example Sywell , were returned to civilian use . Surplus military airfields were closed , and in some cases , for example Beccles , subsequently re @-@ opened as civil aerodromes . The Ministry of Civil Aviation was created to regulate all civil aviation in the UK , and this task remained the responsibility of government departments until the establishment of the independent CAA in 1972 . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n With an expanded infrastructure in place , GA became established after the war when manufacturers such as Cessna and Piper introduced light aircraft designed for the private market . The Cessna 172 , developed from the late 1940s Cessna 170 , was introduced in 1956 , and became the world 's best selling single @-@ engine aeroplane . Single piston @-@ engine aircraft are still the most common class of aircraft in the UK GA fleet . The development of the Rogallo wing in the 1950s fostered the development of hang @-@ gliding during the 1960s and 1970s . The 1960s also saw experiments with motorised hang gliders , but it was not until the 1970s that this blend of technologies started to mature , resulting in the birth of the microlight movement . Another milestone in the development of GA was the 1964 introduction of the Learjet 23 . Although it was not the first business jet , it popularised corporate aviation , and established the personal jet as a \" whole new class of aircraft \" . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n The GA sector operates a range of aircraft , including balloons and airships , gliders , hang gliders , paragliders , microlights , gyrocopters , helicopters , amateur built and mass @-@ produced light aircraft , ex @-@ military aircraft , and business jets . Flights can be broadly categorised as public transport , aerial work , and private flying , the first two of which are commercial activities . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Commercial operations are remunerated activities which fall within the ICAO definition of CAT . Some are , however , closely aligned to , and considered part of , the GA sector . Public transport operations are non @-@ scheduled , on @-@ demand services flying between points specified by the customer , providing a more flexible service than airline travel . Air taxi operations offer charter services for third parties , and business or corporate aviation uses company @-@ owned aircraft to transport employees and clients . Aircraft used in these operations include business jets , helicopters , and twin piston @-@ engine aeroplanes carrying between six and ten people . An example of this type of operation is the transport by helicopter of spectators to the British Formula One grand prix at Silverstone . This involves so many flights that , according to Cranfield Aviation Services , on race day the heliport is temporarily the world 's busiest airport . Aerial work is a small but important component of the commercial GA sector , characterised in its simplest form as remunerated non @-@ transport activities , such as surveying , crop spraying , and emergency services work ( air ambulance and police ) . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Flying schools are commercial businesses engaged in the training of pilots , both for recreational purposes and for those intending to fly professionally . They make widespread use of fixed @-@ wing light aircraft associated with traditional GA , not only for flying lessons but also as club aircraft rented out to qualified pilots for recreational flights . School @-@ owned aircraft account for a significant amount of GA activity , both in terms of hours flown and aircraft movements . The pilot training element is regarded by the GA community as a key benefit that is critical to the supply of pilots for the airline industry . It is claimed by the General Aviation Awareness Council that 60 – 70 per cent of professional pilots have self @-@ financed their flight training at GA schools , and one UK airline operator has stated that the industry must rely on 70 – 80 per cent of new pilots coming from the GA sector . The CAA estimates that between 1996 and 2006 the number of new professional pilots following the unsponsored training route rose from 48 per cent to 59 per cent . The counter argument to this claim is that pilots can be trained outside of the UK , and that the airline industry is not therefore dependent on a healthy GA sector in the UK for its supply of pilots . The CAA concludes that a severe reduction in GA would give \" some merit to the argument that pilot recruitment would be threatened \" , but that the data on flying hours \" does not support such a gloomy outlook . \" Of course , reliance on other countries for pilot training means that the UK foregoes the economic benefit of the training activity . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Private flying can be for both recreational purposes and personal transport , using aircraft that are owned individually , collectively as part of a syndicate , or rented from a flying club . A survey of pilots conducted between 2001 and 2002 indicated that the most common purposes of recreational flights were local flights near the base aerodrome , visits to other aerodromes , and day trips away . Half of all flights landed at the same aerodrome they departed from , and only 9 per cent involved an overnight stay away from home . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Private flying is most associated with the traditional form of factory @-@ produced two and four @-@ seater , single piston @-@ engine training and touring aircraft . Examples of these are the Cessna 152 , Cessna 172 , and Piper PA28 Cherokee , all with their origins in the 1950s , and the more modern designs of Cirrus . The average cost per hour to fly such aircraft has been estimated to be £ 133 , compared to an estimated £ 77 per hour for gliders , and a reported £ 35 per hour for microlights . Recent trends have seen an increase in the use of microlights , and also in recreational helicopter flying following the introduction of smaller and cheaper machines such as the Robinson R22 and R44 . Another growth area in private flying in recent years has been in the use of amateur built aircraft , such as the Van 's Aircraft RV @-@ 4 and the Europa . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n There is a strong vintage aircraft movement in the UK , with two @-@ thirds of the 500 registered historic aircraft active . These cover the whole spectrum of civil and military aviation , examples being the de Havilland Dragon Rapide airliner of the 1930s , and the World War II ( WWII ) Spitfire fighter . There are many post @-@ WWII aircraft which could also be considered historic under a looser definition , including for example 60 ex @-@ military jets such as the Hawker Hunter . Historic aircraft are regular exhibits at air displays , which are claimed to be the second most popular spectator activity after football in the UK . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Competitive gliding in the UK takes place between May and September . Regionals are local competitions , organised and run by one of the bigger gliding clubs in the region , and represent the entry level to glider racing . Races are handicapped according to glider performance , and normally take place over nine days . Success in the regionals allows pilots to progress to the nationals , where there are five classes of competition . These are based on glider performance , the lowest being club class , and then progressing through standard ( maximum 15 metres ( 49 ft ) wingspan , and flaps not permitted ) , 15 metres ( 49 ft ) ( as standard , but flaps are permitted ) , 18 metres ( 59 ft ) ( maximum 18 metres ( 59 ft ) wingspan ) , and finally open @-@ class ( no restrictions ) . Success at national level can lead to a place in the national team and competition at international level . In 2007 the British gliding team was ranked number one , and British pilots took two women 's world championships and the open class European championship . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Handicapped air racing is open to any propeller @-@ driven aircraft capable of maintaining a minimum speed of 100 miles ( 160 km ) per hour in level flight . Races are a case of \" fly low , fly fast , turn left \" , consisting of 4 – 5 laps round a 20 – 25 mile ( 32 – 40 km ) circuit . Faster aircraft are handicapped by starting after slower aircraft , the intention being that the race concludes with all aircraft diving for the finish line together . There are up to 16 races per year , conducted at airfields in the UK , France and the Channel Islands , for prizes that include the Schneider Trophy and King 's Cup , and the season culminates with the British Air Racing and European Air Racing Championships . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": "<|start-PI|>Handicapped air racing is open to any propeller @-@ driven aircraft capable of maintaining a minimum speed of<|end-PI|> <|retrieval|> <|start-KFI|>100 miles ( 160 km ) per hour in level flight<|end-KFI|>. <|start-PI|>Races are a case of \" fly low , fly fast , turn left \" , consisting of<|end-PI|> <|retrieval|> <|start-KFI|>4 – 5 laps round a 20 – 25 mile ( 32 – 40 km ) circuit<|end-KFI|>. Faster aircraft are handicapped by starting after slower aircraft , the intention being that the race concludes with all aircraft diving for the finish line together . <|start-PI|>There are up to 16 races per year , conducted at airfields in<|end-PI|> <|retrieval|> <|start-KFI|>the UK , France and the Channel Islands<|end-KFI|> , <|start-PI|>for prizes that include<|end-PI|> <|retrieval|> <|start-KFI|>the Schneider Trophy and King 's Cup<|end-KFI|> , and the season culminates with the British Air Racing and European Air Racing Championships .", "none": null, "results": [ { "Explanation": "The PI introduces the concept of handicapped air racing and sets up a query about the minimum speed requirement, while the KFI provides the specific speed information. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.", "Ideal Knowledge": "A detailed explanation of handicapped air racing rules and regulations, specifically focusing on the minimum speed requirements for participating aircraft.", "Key Factual Information": "100 miles ( 160 km ) per hour in level flight", "Prefix reformulated": "What is the minimum speed required for aircraft in handicapped air racing?", "Prefix-intent": "Handicapped air racing is open to any propeller @-@ driven aircraft capable of maintaining a minimum speed of", "Verification": "The KFI directly provides the minimum speed requirement mentioned in the PI." }, { "Explanation": "The PI describes the general nature of the races and sets up a query about their structure, while the KFI provides specific details about the number of laps and circuit length. The <|retrieval|> token is inserted to indicate where additional information about race structure might be needed.", "Ideal Knowledge": "A comprehensive guide to handicapped air racing, including detailed information about race formats, course layouts, and typical race distances.", "Key Factual Information": "4 – 5 laps round a 20 – 25 mile ( 32 – 40 km ) circuit", "Prefix reformulated": "What is the structure of handicapped air races?", "Prefix-intent": "Races are a case of \" fly low , fly fast , turn left \" , consisting of", "Verification": "The KFI directly describes the race structure as prompted by the PI." }, { "Explanation": "The PI sets up a query about the locations of the races, and the KFI provides the specific countries and regions. The <|retrieval|> token is inserted to indicate where additional information about race locations might be needed.", "Ideal Knowledge": "A list of airfields and locations where handicapped air races are held, including specific venues in the UK, France, and the Channel Islands, along with any notable features of these locations.", "Key Factual Information": "the UK , France and the Channel Islands", "Prefix reformulated": "Where are handicapped air races conducted?", "Prefix-intent": "There are up to 16 races per year , conducted at airfields in", "Verification": "The KFI directly provides the locations of the races as prompted by the PI." }, { "Explanation": "The PI introduces the concept of prizes for the races and sets up a query about what they are, while the KFI provides specific examples of prestigious trophies. The <|retrieval|> token is inserted to indicate where additional information about race prizes might be needed.", "Ideal Knowledge": "A comprehensive list of prizes and trophies awarded in handicapped air racing, including the history and significance of the Schneider Trophy and King's Cup, as well as any other notable awards.", "Key Factual Information": "the Schneider Trophy and King 's Cup", "Prefix reformulated": "What are the prizes for handicapped air races?", "Prefix-intent": "There are up to 16 races per year , conducted at airfields in the UK , France and the Channel Islands , for prizes that include", "Verification": "The KFI directly lists some of the prizes as prompted by the PI." } ] }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Aerobatic competitions take place for both powered aircraft and gliders , with up to 30 events each year in the UK and Ireland . Starting at the Beginner level , pilots can move up to Standard ( powered aircraft ) or Sports ( glider ) levels , and then on to Intermediate , Advanced , and finally Unlimited classes . Each step up requires a wider repertoire of aerobatic figures and progressively more performance from the aircraft . National championships are awarded annually at Standard / Sports , Intermediate , Advanced ( powered aircraft only ) , and Unlimited levels , and pilots who have reached Advanced and Unlimited levels are eligible for selection to represent the UK in international competition . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": "<|start-PI|>Aerobatic competitions take place for both powered aircraft and gliders , with up to<|end-PI|> <|retrieval|> <|start-KFI|>30 events each year in the UK and Ireland<|end-KFI|> . <|start-PI|>Starting at the Beginner level , pilots can move up to<|end-PI|> <|retrieval|> <|start-KFI|>Standard ( powered aircraft ) or Sports ( glider ) levels , and then on to Intermediate , Advanced , and finally Unlimited classes<|end-KFI|> . Each step up requires a wider repertoire of aerobatic figures and progressively more performance from the aircraft . <|start-PI|>National championships are awarded annually at<|end-PI|> <|retrieval|> <|start-KFI|>Standard / Sports , Intermediate , Advanced ( powered aircraft only ) , and Unlimited levels<|end-KFI|> , and <|start-PI|>pilots who have reached Advanced and Unlimited levels are eligible for<|end-PI|> <|retrieval|> <|start-KFI|>selection to represent the UK in international competition<|end-KFI|> .", "none": null, "results": [ { "Explanation": "The PI introduces aerobatic competitions and sets up a query about their frequency, while the KFI provides the specific number and location. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.", "Ideal Knowledge": "A comprehensive list of aerobatic competitions for powered aircraft and gliders in the UK and Ireland, including the total number of events held annually.", "Key Factual Information": "30 events each year in the UK and Ireland", "Prefix reformulated": "How many aerobatic competitions take place for powered aircraft and gliders in the UK and Ireland annually?", "Prefix-intent": "Aerobatic competitions take place for both powered aircraft and gliders , with up to", "Verification": "The KFI directly provides the number of events and location information requested in the PI." }, { "Explanation": "The PI introduces the concept of progression in aerobatic competitions, and the KFI provides the specific levels. The <|retrieval|> token is inserted to indicate where additional information about the competition levels might be needed.", "Ideal Knowledge": "A detailed explanation of the different levels in aerobatic competitions for both powered aircraft and gliders, including the specific requirements and skills needed for each level.", "Key Factual Information": "Standard ( powered aircraft ) or Sports ( glider ) levels , and then on to Intermediate , Advanced , and finally Unlimited classes", "Prefix reformulated": "What are the levels pilots can progress through in aerobatic competitions?", "Prefix-intent": "Starting at the Beginner level , pilots can move up to", "Verification": "The KFI directly lists the progression of levels for pilots in aerobatic competitions, as prompted by the PI." }, { "Explanation": "The PI sets up a query about the levels of national championships, and the KFI provides the specific levels. The <|retrieval|> token is inserted to indicate where additional information about national championships might be needed.", "Ideal Knowledge": "Information on the structure of national aerobatic championships, including the specific levels at which they are awarded for both powered aircraft and gliders.", "Key Factual Information": "Standard / Sports , Intermediate , Advanced ( powered aircraft only ) , and Unlimited levels", "Prefix reformulated": "At which levels are national championships awarded annually in aerobatic competitions?", "Prefix-intent": "National championships are awarded annually at", "Verification": "The KFI directly lists the levels at which national championships are awarded, as prompted by the PI." }, { "Explanation": "The PI introduces the concept of eligibility for pilots at higher levels, and the KFI provides the specific opportunity they are eligible for. The <|retrieval|> token is inserted to indicate where additional information about international competition selection might be needed.", "Ideal Knowledge": "Details on the selection process for international aerobatic competitions, including the requirements for pilots to be considered for representing their country.", "Key Factual Information": "selection to represent the UK in international competition", "Prefix reformulated": "What are pilots who have reached Advanced and Unlimited levels eligible for in aerobatic competitions?", "Prefix-intent": "pilots who have reached Advanced and Unlimited levels are eligible for", "Verification": "The KFI directly states what pilots at Advanced and Unlimited levels are eligible for, as prompted by the PI." } ] }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Parachute competitions are held at club , regional , national and international levels , and include the disciplines of accuracy landings , freefall gymnastics , formation skydiving , canopy formation , freestyle and freeflying , and skysurfing . British teams consistently win medals in canopy formation world championships , and a British team took the 2006 world championship in women 's 4 @-@ way formation skydiving . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Aerodrome is a collective term for any location from which flying operations take place , although more specific terminology can be used to characterise its purpose . The CAA strategic review of GA applies the term airport to locations which predominantly support large scale commercial operations , and airfield to locations which predominantly support GA operations . The General Aviation Small Aerodrome Research Study ( GASAR ) analysed 687 aerodromes in England which come under the scope of GA , classifying 374 into six types . These range in size from regional airports to the smallest farm strip , although 84 per cent of GA flights operate from 134 of the larger aerodromes in the first four categories . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n The factors used in determining how an individual aerodrome is categorised by the GASAR study are based broadly on size and facilities . The six types of aerodrome are described , in size order , as : regional airports ( e.g. East Midlands ) ; major GA airports ( e.g. Oxford ) ; developed GA airfields ( e.g. Andrewsfield ) ; basic GA airfields ( e.g. Rufforth ) ; developed airstrips ( e.g. Tilstock ) ; and basic airstrips ( e.g. Chilbolton in Hampshire ) . The actual criteria used to categorise aerodromes were complex , using 28 different parameters , backed up with a peer review by experienced GA pilots . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Airports generally have long , fully lit , hard @-@ surfaced runways , full air traffic control , and navigation and landing aids . They are usually located on urban fringes , support commercial and business operations , and often exclude certain types of light aircraft . At the more rurally located airfields , the lighter end of aviation , such as microlight and gliding activities , becomes increasingly prevalent , and there are few or no commercial operations other than flying schools . At this level runways are generally shorter , and grass surfaces are increasingly common . Navigation aids are increasingly scarce , being more basic where they are available , and informal ground to air radio communication replaces air traffic control . The smallest airfields are too small to feature on general purpose Ordnance Survey ( OS ) maps , and lack basic facilities such as fuel and maintenance . The majority of airstrips are basically single short grass runways with no supporting facilities , although the presence of a hangar is not uncommon at the larger examples . They do not feature on OS maps , and are owned by private clubs or , more commonly , individuals . \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": null, "none": "", "results": null }

[ { "content": "You are an AI assistant tasked with analyzing a passage of text to insert retrieval tokens and identify Prefix-intent (PI) and Key Factual Information (KFI). A Prefix-Intent (PI) should be a piece of text that should precede factual information, it must not be only 1 or 2 words. Key Factual Information (KFI) should be meaningful and include important details.\n\nYour goal is to insert a <|retrieval|> token after it is clear what the intent of the text is (e.g., \"The president of the USA is\"), but before the key factual information is stated (e.g., \"Joe Biden\"). Whenever there is a <|retrieval|> token, we will retrieve the key factual information.\n\nHere is the passage you will be working with:\n\n<passage>\n Most aerodromes used for public transport operations are required to be licensed by the CAA . To be granted a licence an aerodrome operator must satisfy the CAA that : the physical conditions at the aerodrome , and its environs , are acceptable ; the scale of equipment , and facilities provided , are adequate for the flying activities which are expected to take place ; an effective safety management system is in place ; and that staff are competent and , where necessary , suitably qualified . Aerodromes classified as developed GA airfields or larger by the GASAR study are , with few exceptions , licensed . Only two basic GA airfields , Silverstone and Duxford , are licensed , and all airstrips are unlicensed . The Light Aviation Airports Study Group , a joint CAA @-@ industry initiative , was established in 2005 to review the regulation of light aviation aerodromes . A particular focus of this group was a review of the restrictions placed on unlicensed aerodromes . The group concluded that the requirement for public transport operations to be conducted only from licensed aerodromes should be further reviewed in the context of corresponding international and European requirements . It also recommended that restrictions on flight training at unlicensed aerodromes should be lifted , and this was permitted from April 2010 \n\n</passage>\n\nFollow these steps to complete the task:\n\n1. Inserting retrieval tokens:\n a. Read through the text carefully.\n b. Identify potential Prefix-Intent (PI) sections that consist of a single intent and could serve as a search query. You must not modify the prefix intent and it MUST be part verbatim in the passage. It cannot be single or only a few words.\n c. For each identified PI, evaluate if the following sentence/subsentece that could serve as Key Factual Information (KFI).\n d. Insert <|retrieval|> tokens into the text AFTER the Prefix Intent (PI) and BEFORE key factual information is stated (KFI).\n e. Key Factual information should never intersect with Prefix Intent. Ensure this!\n\n2. Identifying Prefix-Intent (PI) and Key Factual Information (KFI):\n a. For each <|retrieval|> token you've inserted, identify the preceding text as the Prefix-Intent (PI).\n b. Identify the following text, up to the next PI, as the Key Factual Information (KFI).\n c. Ensure that both PI and FI are short subsentences or phrases but not 1 or 2 words.\n\n3. Insert <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|>.\n a. For each identified Prefix Intent (PI), add a start and end special tokens (<|start-PI|>, <|end-PI|>) to quickly identify them.\n b. For each identified Key Factual Information (KFI), also add a start and end special tokens (<|start-KFI|>, <|end-KFI|>) \n c. Ensure that Prefix Intent (PI) and Key Factual Information (KFI) special tokens are inserted.\n\n4. Describing ideal retrieved information:\n a. Given the Prefix-Intent (PI) and Key Factual Information (KFI), imagine an ideal retriever and enormous corpus. \n b. Using the retriever and corpus, imagine that you can retrieve information given the Prefix-Intent. \n c. Describe how that content should look like so it helps generating the Key Factual Information (KFI)\n\n\n5. Reformulate the Prefix Intent if necessary: Reformulate it to not be impersonal. This means that if the subject it is not identified (e.g. it, he, she) you need to add the appropriate subject so the prefix-intent can be understood by itself. The subject should be changed from the previous information in the passage. Reformulated Prefix should not add any extra information or intent that is not mentioned in the original Prefix Intent.\n\n\n6. Output format:\n For each pair of Prefix-Intent (PI) and Key Factual Information (KFI) you identify, provide the following output separating each example in a JSON list output. Ensure that it is actually a JSON.\n\n - Prefix-intent: [Write the Prefix Intent here]\n - Prefix reformulated: [Write the Prefix Intent reformulated]\n - Key Factual Information: [Write the Key Factual Information here]\n - Verification: [Confirm that the KFI can be directly derived from the PI given a strictly connected context]\n - Ideal Knowledge: [Describe how the ideal retrieved content should look to generate the KFI]\n - Explanation: [Briefly explain how the PI and KFI connect and why you chose to insert the <|retrieval|> token in that position]\n\n\n7. Final step:\n After you have completed the analysis and provided outputs for all identified PI-KFI pairs, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions and also <|start-PI|>,<|end-PI|>, <|start-KFI|>, <|end-KFI|> special tokens. Ensure that PI and KFI special tokens are inside the updated passage.\n\nHere's an example of the expected output format. Updated Passage is constructed using PI-KFI pairs plus the additional special tokens:\n\nPassage: Girmal Falls General This waterfall extends to a height of up to 100 feet, making it the highest waterfall of Gujarat. The picturesque beauty of this site makes it popular among visitors and people of the region alike. The water swiftly falls from a great height, creating a fog like condition that’s eye catching. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. The fall comes to its best form at the time of monsoon and provides an immensely striking appearance. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\nOutput:\n{ \"results\" : [\n {\n \"Prefix-intent\": \"Girmal Falls General This waterfall extends to a height of up to\",\n \"Prefix reformulated\": \"What is the height of Girmal Falls?\",\n \"Key Factual Information\": \"100 feet, making it the highest waterfall of Gujarat\",\n \"Verification\": \"The KFI directly provides the height information requested in the PI.\",\n \"Ideal Knowledge\": \"A list of waterfalls in Gujarat, India, with their heights and rankings, specifically mentioning Girmal Falls as the highest at 100 feet.\",\n \"Explanation\": \"The PI introduces Girmal Falls and sets up a query about its height, while the KFI provides the specific measurement and its significance. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the factual statement.\"\n },\n {\n \"Prefix-intent\": \"The water swiftly falls from a great height, creating\",\n \"Prefix reformulated\": \"What does the water falling from Girmal Falls create?\",\n \"Key Factual Information\": \"a fog like condition that's eye catching\",\n \"Verification\": \"The KFI directly describes the effect created by the falling water, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Detailed descriptions of the visual effects created by high waterfalls, including the formation of mist or fog-like conditions due to water droplets in the air.\",\n \"Explanation\": \"The PI describes the water falling and sets up a query about its effect, while the KFI provides the specific visual phenomenon created. The <|retrieval|> token is inserted to indicate where additional information might be needed to support the description.\"\n },\n {\n \"Prefix-intent\": \"The fall comes to its best form at the time of\",\n \"Prefix reformulated\": \"When does Girmal Falls come to its best form?\",\n \"Key Factual Information\": \"monsoon and provides an immensely striking appearance\",\n \"Verification\": \"The KFI directly answers when the waterfall is at its best and describes its appearance, as prompted by the PI.\",\n \"Ideal Knowledge\": \"Information on the seasonal variations of Girmal Falls, highlighting the monsoon season's impact on water flow and visual appeal.\",\n \"Explanation\": \"The PI sets up a query about the best time to view the falls, and the KFI provides the specific season and describes the appearance. The <|retrieval|> token is inserted to indicate where additional information about seasonal changes might be needed.\"\n }],\n\"Updated_Passage\":\"<|start-PI|> Girmal Falls General This waterfall extends to a height of up to <|end-PI|> <|retrieval|> <|start-KFI|> 100 feet, making it the highest waterfall of Gujarat <|end-KFI|>. The picturesque beauty of this site makes it popular among visitors and people of the region alike. <|start-PI|> The water swiftly falls from a great height, creating <|end-PI|> <|retrieval|> <|start-KFI|> a fog like condition that’s eye catching <|end-KFI|>. The government of this state is working on many projects to make this place an ideal picnic spot and a tourist attraction. <|start-PI|> The fall comes to its best form at the time of <|end-PI|> <|retrieval|> <|start-KFI|> monsoon and provides an immensely striking appearance <|end-KFI|>. Some of the best natural features of Gujarat can be explored in this place. This place is a nice and refreshing retreat for any traveler.\"\n}\n\n \nRemember to analyze the entire passage, identify all relevant PI-KFI pairs, and provide the output for each pair as specified. After completing the analysis, present the original passage with the <|retrieval|> tokens inserted in their appropriate positions as part of a new key in the JSON file. Only Ouput the JSON File.", "role": "user" } ]

{ "Updated_Passage": "Most aerodromes used for public transport operations are required to be licensed by the CAA . <|start-PI|> To be granted a licence an aerodrome operator must satisfy the CAA that <|end-PI|> <|retrieval|> <|start-KFI|> the physical conditions at the aerodrome , and its environs , are acceptable ; the scale of equipment , and facilities provided , are adequate for the flying activities which are expected to take place ; an effective safety management system is in place ; and that staff are competent and , where necessary , suitably qualified <|end-KFI|> . <|start-PI|> Aerodromes classified as developed GA airfields or larger by the GASAR study are <|end-PI|> <|retrieval|> <|start-KFI|> with few exceptions , licensed <|end-KFI|> . Only two basic GA airfields , Silverstone and Duxford , are licensed , and all airstrips are unlicensed . <|start-PI|> The Light Aviation Airports Study Group , a joint CAA @-@ industry initiative , was established in 2005 to <|end-PI|> <|retrieval|> <|start-KFI|> review the regulation of light aviation aerodromes <|end-KFI|> . A particular focus of this group was a review of the restrictions placed on unlicensed aerodromes . <|start-PI|> The group concluded that the requirement for public transport operations to be conducted only from licensed aerodromes should be <|end-PI|> <|retrieval|> <|start-KFI|> further reviewed in the context of corresponding international and European requirements <|end-KFI|> . <|start-PI|> It also recommended that restrictions on flight training at unlicensed aerodromes should be <|end-PI|> <|retrieval|> <|start-KFI|> lifted , and this was permitted from April 2010 <|end-KFI|>", "none": null, "results": [ { "Explanation": "The PI sets up a query about the requirements for obtaining an aerodrome licence, and the KFI provides a detailed list of these requirements. The <|retrieval|> token is inserted to indicate where additional information about licensing requirements might be needed.", "Ideal Knowledge": "A comprehensive list of CAA requirements for aerodrome licensing, including details on physical conditions, equipment, safety management systems, and staff qualifications.", "Key Factual Information": "the physical conditions at the aerodrome , and its environs , are acceptable ; the scale of equipment , and facilities provided , are adequate for the flying activities which are expected to take place ; an effective safety management system is in place ; and that staff are competent and , where necessary , suitably qualified", "Prefix reformulated": "What requirements must an aerodrome operator satisfy to be granted a licence by the CAA?", "Prefix-intent": "To be granted a licence an aerodrome operator must satisfy the CAA that", "Verification": "The KFI directly lists the requirements that aerodrome operators must meet to obtain a licence from the CAA, as prompted by the PI." }, { "Explanation": "The PI introduces a specific classification of aerodromes, and the KFI provides information about their licensing status. The <|retrieval|> token is inserted to indicate where additional information about aerodrome classifications and licensing might be needed.", "Ideal Knowledge": "A detailed breakdown of aerodrome classifications according to the GASAR study, with specific information on licensing requirements and exceptions for each category.", "Key Factual Information": "with few exceptions , licensed", "Prefix reformulated": "What is the licensing status of aerodromes classified as developed GA airfields or larger by the GASAR study?", "Prefix-intent": "Aerodromes classified as developed GA airfields or larger by the GASAR study are", "Verification": "The KFI directly states the licensing status of the aerodromes mentioned in the PI." }, { "Explanation": "The PI introduces the study group and sets up a query about its purpose, while the KFI provides the specific reason for its establishment. The <|retrieval|> token is inserted to indicate where additional information about the study group's objectives might be needed.", "Ideal Knowledge": "Detailed information about the Light Aviation Airports Study Group, including its formation, objectives, and key areas of focus in reviewing light aviation aerodrome regulations.", "Key Factual Information": "review the regulation of light aviation aerodromes", "Prefix reformulated": "What was the purpose of establishing the Light Aviation Airports Study Group in 2005?", "Prefix-intent": "The Light Aviation Airports Study Group , a joint CAA @-@ industry initiative , was established in 2005 to", "Verification": "The KFI directly states the purpose of establishing the study group mentioned in the PI." }, { "Explanation": "The PI sets up a specific conclusion of the study group, and the KFI provides the details of that conclusion. The <|retrieval|> token is inserted to indicate where additional information about the group's findings and recommendations might be needed.", "Ideal Knowledge": "A summary of the Light Aviation Airports Study Group's findings, focusing on their recommendations for reviewing public transport operation requirements in light of international and European standards.", "Key Factual Information": "further reviewed in the context of corresponding international and European requirements", "Prefix reformulated": "What did the Light Aviation Airports Study Group conclude about the requirement for public transport operations to be conducted only from licensed aerodromes?", "Prefix-intent": "The group concluded that the requirement for public transport operations to be conducted only from licensed aerodromes should be", "Verification": "The KFI directly states the conclusion of the study group regarding the requirement mentioned in the PI." }, { "Explanation": "The PI introduces a recommendation by the study group, and the KFI provides the specific recommendation and its implementation date. The <|retrieval|> token is inserted to indicate where additional information about changes in flight training regulations might be needed.", "Ideal Knowledge": "Detailed information about the changes in regulations for flight training at unlicensed aerodromes, including the specific recommendations made by the Light Aviation Airports Study Group and the date of implementation.", "Key Factual Information": "lifted , and this was permitted from April 2010", "Prefix reformulated": "What did the Light Aviation Airports Study Group recommend regarding restrictions on flight training at unlicensed aerodromes?", "Prefix-intent": "It also recommended that restrictions on flight training at unlicensed aerodromes should be", "Verification": "The KFI directly states the recommendation and its implementation, as prompted by the PI." } ] }