| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| GIM | Comprehensiveness76 | 33 | 1mo ago | ||
| SciFact | AttnLRP | Comprehensiveness74 | 33 | 1mo ago | |
| Movie | GIM | Comprehensiveness83 | 33 | 1mo ago | |
| HateXplain | AttnLRP | Comprehensiveness85 | 33 | 1mo ago | |
| FEVER | AttnLRP | Comprehensiveness0.75 | 33 | 1mo ago | |
| BoolQ | GIM | Comprehensiveness72 | 33 | 1mo ago | |
| ImageNet (test) | SIG | GAP0.3104 | 30 | 2mo ago | |
| ImageNet 1000 samples | FAMPE | Insertion Score57.44 | 28 | 2mo ago | |
| Image data 224 x 224 | h-SHAP | Avg Execution Time (s)0.24 | 28 | 4mo ago | |
| Synthetic VAR Data | WI | Kendall's τ0.94 | 25 | 26d ago | |
| Mini-ImageNet 500 randomly sampled images (val) | DiffIG | DiffID Score42.98 | 24 | 1mo ago | |
| Oxford-IIIT Pet full (val) | DiffIG | DiffID63.68 | 24 | 1mo ago | |
| ImageNet 1000 samples (val) | GuidedGradCam | Infidelity4.37 | 24 | 2mo ago | |
| CUB | H-Sets | ROAD_AOPC0.65 | 24 | 2mo ago | |
| ImageNet | H-Sets | ROAD_AOPC38 | 24 | 2mo ago | |
| SST2 | Grad*Inp | LO-0.199 | 18 | 4mo ago | |
| IMDB (test) | Grad*Inp | LO-0.0892 | 18 | 4mo ago | |
| Rotten Tomatoes fine-tuned | Grad*Inp | LO-0.121 | 18 | 4mo ago | |
| IRIS | SHAP | INFD0.007 | 17 | 4mo ago | |
| ImageNet and COCO | Expected Gradients | AUC of Deletion13.2 | 14 | 1mo ago | |
| MS-CXR text (test) | RISE | Conf. Drop (%)1.16 | 13 | 4mo ago | |
| Rotten Tomatoes | SHAP | INFD0 | 13 | 4mo ago | |
| CIFAR10 | SHAP | INFD0.003 | 13 | 4mo ago | |
| FMNIST | SHAP | INFD0 | 13 | 4mo ago | |
| MEPS | SHAP | INFD0 | 12 | 4mo ago |
