Research map / Label-efficient and robust learning / Recognition and long-tail
What You See is What You Classify: Black Box Attributions
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 learned class-specific Explainer masks instead of the strongest listed baseline | VGG-16 · VOC-2007 | Acc: 59.63 -> 74.60 | Helped |
| Use learned class-specific Explainer masks instead of the strongest listed baseline | VGG-16 · VOC-2007 | IoU: 39.94 -> 38.04 | Hurt |
| Use learned class-specific Explainer masks instead of Grad-CAM | ResNet-50 · VOC-2007 | Acc: 68.54 -> 73.56 | Helped |
| Use learned class-specific Explainer masks instead of the strongest listed baseline | ResNet-50 · VOC-2007 | IoU: 36.46 -> 34.23 | Hurt |
| Use learned class-specific Explainer masks instead of Grad-CAM | VGG-16 · COCO-2014 | Acc: 68.70 -> 69.90 | Helped |
| Use learned class-specific Explainer masks instead of the strongest listed baseline | VGG-16 · COCO-2014 | IoU: 31.16 -> 31.59 | Helped |
| Use learned class-specific Explainer masks instead of the strongest listed baseline | ResNet-50 · COCO-2014 | Acc: 69.57 -> 65.30 | Hurt |
| Use learned class-specific Explainer masks instead of the strongest listed baseline | ResNet-50 · COCO-2014 | IoU: 30.76 -> 27.42 | Hurt |
| Generate masks with a trained Explainer rather than repeated RISE perturbations | VGG-16 attribution inference · VOC-2007 | inference time: 01:15:29 -> 00:00:08 | Helped |
| Generate masks with a trained Explainer rather than repeated RISE perturbations | VGG-16 attribution inference · COCO-2014 | inference time: 06:00:47 -> 00:00:34 | Helped |
Closest papers by meaning
- On Gradient-like Explanation under a Black-box Setting: When Black-box Explanations Become as Good as White-box (ICML 2024 · similarity 0.87)
- How to Probe: Simple Yet Effective Techniques for Improving Post-hoc Explanations (ICLR 2025 · similarity 0.87)
- GEFA: A General Feature Attribution Framework Using Proxy Gradient Estimation (ICML 2025 · similarity 0.87)
- Benchmarking the Attribution Quality of Vision Models (NeurIPS 2024 · similarity 0.86)
- One Wave To Explain Them All: A Unifying Perspective On Feature Attribution (ICML 2025 · similarity 0.86)
- Pixel-level Certified Explanations via Randomized Smoothing (ICML 2025 · similarity 0.85)
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.