attention-rollout / description.md
wzuidema
Update description.md
28caaa6 unverified
|
raw
history blame
623 Bytes

Attention Rollout -- RoBERTa

In this demo, we use the RoBERTa language model (optimized for masked language modelling and finetuned for sentiment analysis). The model predicts for a given sentences whether it expresses a positive, negative or neutral sentiment. But how does it arrive at its classification? This is, surprisingly perhaps, very difficult to determine.

Abnar & Zuidema (2020) proposed a method for Transformers called "Attention Rollout", which was further refined by Chefer et al. (2021) into Gradient-weighted Rollout. Here we compare it to another popular method called Integrated Gradients.