Generic Attention-model Explainability for Interpreting Bi-Modal and Encoder-Decoder Transformers.

Hila Chefer, Shir Gur, Lior Wolf

Browse the full ICCV paper archive.