Research map / Label-efficient and robust learning / Recognition and long-tail
Distribution-Based Feature Attribution for Explaining the Predictions of Any Classifier
Xinpeng Li, Kai Ming Ting
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 changed | Where | Measured | Effect |
|---|---|---|---|
| use DFAX distribution-based attribution instead of DLIME | feature attribution · Average over 10 datasets | Deletion score: .4595 -> .3244 | Improved |
| use DFAX instead of DLIME | feature attribution · Average over 10 datasets | Insertion score: .6612 -> .7708 | Improved |
| use DFAX instead of DLIME | feature attribution · Diabetes | Insertion score: .8781 vs .8738 | No clear effect |
| use GKDE instead of SiNNE in DFAX (ablation) | DFAX · Average over 10 datasets | Insertion score: .7470 -> .7708 | Improved |
| use DFAX instead of DLIME for HER2st gene selection | HER2st classifier · HER2st | Accuracy after masking: 79.51% -> 95.64% | Improved |
| SHAP and MAPLE as attribution methods vs random | feature attribution · Average over 10 datasets | Insertion score: SHAP .5463, MAPLE .5800 vs Random .5838 | Hurt |
Closest papers by meaning
- GEFA: A General Feature Attribution Framework Using Proxy Gradient Estimation (ICML 2025 · similarity 0.90)
- Provably Better Explanations with Optimized Aggregation of Feature Attributions (ICML 2024 · similarity 0.89)
- Feature Attribution with Necessity and Sufficiency via Dual-stage Perturbation Test for Causal Explanation (ICML 2024 · similarity 0.87)
- On Gradient-like Explanation under a Black-box Setting: When Black-box Explanations Become as Good as White-box (ICML 2024 · similarity 0.86)
- Selective Explanations (NeurIPS 2024 · similarity 0.86)
- Prospector Heads: Generalized Feature Attribution for Large Models & Data (ICML 2024 · similarity 0.84)
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.
