Research map / Learning theory and algorithms / Prediction and decision losses
A Unified Taylor Framework for Revisiting Attribution Methods
Huiqi Deng, Na Zou, Mengnan Du, Weifu Chen, Guocan Feng, Xia Hu
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 |
|---|---|---|---|
| Include second-order Taylor terms | Taylor explanation · MNIST, C-sg | infidelity: 0.48 -> 0.27 | Improved |
| Include second-order Taylor terms | Taylor explanation · MNIST, C-sp | infidelity: 0.84 -> 0.75 | Improved |
| Use Expected Gradients with multiple Gaussian-noise baselines instead of a single fixed baseline | Integrated Gradients · ImageNet object localization | bounding box accuracy | Improved |
| Follow more of the three attribution principles | Attribution methods · MNIST | infidelity | Improved |
| Follow more of the three attribution principles | Attribution methods · ImageNet | bounding box accuracy | Improved |
| Use a black image as the baseline | Integrated Gradients · ImageNet corgi visualization | feature attribution fidelity | Hurt |
Closest papers by meaning
- Taming Hyperparameter Sensitivity in Data Attribution: Practical Selection Without Costly Retraining (NeurIPS 2025 · similarity 0.88)
- $\texttt{dattri}$: A Library for Efficient Data Attribution (NeurIPS 2024 · similarity 0.87)
- quanda: An Interpretability Toolkit for Training Data Attribution Evaluation (NeurIPS 2026 · similarity 0.87)
- Benchmarking the Attribution Quality of Vision Models (NeurIPS 2024 · similarity 0.87)
- How to Probe: Simple Yet Effective Techniques for Improving Post-hoc Explanations (ICLR 2025 · similarity 0.87)
- Provably Better Explanations with Optimized Aggregation of Feature Attributions (ICML 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.
