Transferring Visual Explainability of Self-Explaining Models through Task Arithmetic

Open in new window