Code for the paper Visual Explanations of Image–Text Representations via Multi-Modal Information Bottleneck Attribution
Primary LanguageJupyter Notebook