Update README.md

README.md

@@ -175,7 +175,7 @@ We trained 2 models on the above dataset:
 - [NovelWriting-Outline-PRM-Qwen2.5-0.5B-Reward](https://huggingface.co/mrzjy/NovelWriting-Outline-PRM-Qwen2.5-0.5B-Reward): The PRM for outline generation task, trained by using TRL library ([Refer to Doc](https://huggingface.co/docs/trl/prm_trainer)).
   - Note: This model is trained with `train_on_last_step_only` flag set to `True`
 
-## 4. Performance Evaluation
+## 4. Usage & Performance Evaluation
 
 ### 4.1 Accuracy Metric
 
@@ -184,13 +184,12 @@ We trained 2 models on the above dataset:
 ```
 ```
 
-### 4.2 LLM Sampling with PRM
+### 4.2 Sequential Rejection Sampling
 
-Without delving into further reinforcement learning or policy updates, can we directly apply PRM with our LLMs? The answer is YES!
+Without delving into further reinforcement learning, can we directly apply PRM with our LLMs? The answer is YES!
 
-#### 4.2.1 Test-Time Scaling
+- Test-Time Scaling
 
-#### 4.2.2 Sequential Rejection Sampling
 
 - Case Study