(작성중....) Transformer Interpretability Beyond Attention Visualization colab: colab.research.google.com/github/hila-chefer/Transformer-Explainability/blob/main/BERT_explainability.ipynb github.com/hila-chefer/Transformer-Explainability Vision transformer에서 relevance score를 계산하는 방법을 제시하는 논문 Method consists of 3 phases: Calculating relevance for each attention matrix using our novel formulation of ..
