Metric: Insertion AUC score (ResNet-101) (higher is better)
| # | Model↕ | Insertion AUC score (ResNet-101)▼ | Extra Data | Paper | Date↕ | Code |
|---|---|---|---|---|---|---|
| 1 | SMDL-Attribution (ICLR version) | 0.7262 | No | Less is More: Fewer Interpretable Region via Sub... | 2024-02-14 | Code |
| 2 | Grad-CAM | 0.7224 | No | Grad-CAM: Visual Explanations from Deep Networks... | 2016-10-07 | Code |
| 3 | RISE | 0.7193 | No | RISE: Randomized Input Sampling for Explanation ... | 2018-06-19 | Code |
| 4 | HSIC-Attribution | 0.6843 | No | Making Sense of Dependence: Efficient Black-box ... | 2022-06-13 | Code |
| 5 | LIME | 0.6812 | No | "Why Should I Trust You?": Explaining the Predic... | 2016-02-16 | Code |
| 6 | Kernel SHAP | 0.6763 | No | A Unified Approach to Interpreting Model Predict... | 2017-05-22 | Code |
| 7 | Saliency | 0.6585 | No | Deep Inside Convolutional Networks: Visualising ... | 2013-12-20 | Code |
| 8 | Integrated Gradients | 0.0422 | No | Axiomatic Attribution for Deep Networks | 2017-03-04 | Code |