Research map / Large language models / AI and society
What I Cannot Predict, I Do Not Understand: A Human-Centered Evaluation Framework for Explainability Methods
Julien Colin, Thomas FEL, Remi Cadene, Thomas Serre
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 |
|---|---|---|---|
| Add Grad-CAM explanations during training | Human meta-predictors · Husky vs. Wolf | Utility: 0.95 -> 1.34 | Improved |
| Add SmoothGrad explanations during training | Human meta-predictors · Husky vs. Wolf | Utility: 1.20 | Improved |
| Add Occlusion explanations during training | Human meta-predictors · Husky vs. Wolf | Utility: 1.22 | Improved |
| Add Saliency explanations during training | Human meta-predictors · Husky vs. Wolf | Utility: 1.06 | No clear effect |
| Add Saliency explanations during training | Human meta-predictors · Leaves | Utility: 1.13 | Improved |
| Add SmoothGrad explanations during training | Human meta-predictors · Leaves | Utility: 1.13 | Improved |
| Add Grad-CAM or Occlusion explanations during training | Human meta-predictors · Leaves | Utility: 1.10 | No clear effect |
| Add any attribution method during training | Human meta-predictors · ImageNet (Kit Fox vs Red Fox) | Utility: 0.90 - 1.00 | No clear effect |
| Replace real explanation with non-informative bottom-up saliency (control) (ablation) | Human meta-predictors · Husky vs. Wolf | Utility: 0.95 | Hurt |
| Replace real explanation with non-informative control (ablation) | Human meta-predictors · ImageNet (Kit Fox vs Red Fox) | session 3 accuracy: 58.8 -> 48.5 | Hurt |
| Use more faithful attribution methods (higher 1 - Deletion) | Attribution methods · All three datasets | Utility | No clear effect |
| Use Insertion and muFidelity instead of Deletion as faithfulness metric | Attribution methods · All three datasets | correlation with Utility | No clear effect |
Closest papers by meaning
- Position: Explainable AI Cannot Advance Without Better User Studies (ICML 2025 · similarity 0.91)
- Towards Unifying Evaluation of Counterfactual Explanations: Leveraging Large Language Models for Human-Centric Assessments (AAAI 2025 · similarity 0.87)
- Utilizing Human Behavior Modeling to Manipulate Explanations in AI-Assisted Decision Making: The Good, the Bad, and the Scary (NeurIPS 2024 · similarity 0.87)
- Navigating the Maze of Explainable AI: A Systematic Approach to Evaluating Methods and Metrics (NeurIPS 2024 · similarity 0.86)
- When Models Know More Than They Can Explain: Quantifying Knowledge Transfer in Human-AI Collaboration (NeurIPS 2025 · similarity 0.86)
- Decoding AI’s Nudge: A Unified Framework to Predict Human Behavior in AI-Assisted Decision Making (AAAI 2024 · similarity 0.86)
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.
