AI Paper Atlas

Research map / Vision-language and multimodal / Human and embodied modeling

Visual Explanations of Image-Text Representations via Multi-Modal Information Bottleneck Attribution

Ying Wang, Tim G. J. Rudner, Andrew Gordon Wilson

NeurIPS 2023 · Vision-language and multimodal › Human and embodied modeling · approach: Language-grounded visual concept learning · 4 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Use M2IB instead of the strongest baseline on MS-CXR image localizationM2IB · MS-CXR image% localization IoU: 21.65 -> 22.59Improved
Use M2IB instead of Saliency on MS-CXR image ROAR+M2IB · MS-CXR image% ROAR+: 25.46 -> 38.7Improved
Use M2IB instead of the transformer-specific method on Conceptual Captions text ROAR+M2IB · Conceptual Captions text% ROAR+: 53.57 -> 60.41Improved
Use M2IB instead of the transformer-specific method on Conceptual Captions image confidence increaseM2IB · Conceptual Captions image% confidence increase: 37.41 -> 41.55Improved
Use M2IB instead of Kernel SHAP on Conceptual Captions text confidence increaseM2IB · Conceptual Captions text% confidence increase: 46.87 -> 38.55Hurt
Use M2IB instead of RISE on MS-CXR text confidence increaseM2IB · MS-CXR text% confidence increase: 57.2 -> 35.48Hurt
Use M2IB instead of RISE on MS-CXR text confidence dropM2IB · MS-CXR text% confidence drop: 1.16 -> 2.28Hurt
Use M2IB instead of RISE on Conceptual Captions image confidence dropM2IB · Conceptual Captions image% confidence drop: 1.12 -> 1.11No clear effect
Use M2IB instead of Saliency on MS-CXR image confidence dropM2IB · MS-CXR image% confidence drop: 0.81 -> 0.55Improved
Choose layer 9, beta=0.1, sigma^2=1 instead of the worst grid setting (ablation)M2IB · Conceptual Captions image% ROAR+: 6.42 -> 12.72Improved
Choose the best versus worst hyperparameters (ablation)M2IB · MS-CXR image% ROAR+: 25.20 -> 37.17Improved
Choose layer 9 and moderate beta for text on Conceptual Captions (ablation)M2IB · Conceptual Captions text% ROAR+: 8.46 -> 18.16Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy