AttnLRP: Attention-Aware Layer-wise Relevance Propagation for Transformers

Open in new window