Feature Importance Assessment on ImageNet-R (test)
19.49Insertion ScoreChefer-LRP
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Chefer-LRPModel=EVA-L-14/196, Training Scheme=EVA2025.07 | 19.49 | 46.54 | |
| CoIBAModel=EVA-L-14/196, Training Scheme=EVA2025.07 | 9.29 | 59.97 | |
| Chefer-LRPModel=ViT*-B-16/224, Training Scheme=CLIP + Massive Regularization2025.07 | 7.72 | 36.7 | |
| Chefer-LRPModel=DeiT3-B-16/224, Training Scheme=DeiT32025.07 | 5.91 | 37.55 | |
| CoIBAModel=ViT*-B-16/224, Training Scheme=CLIP + Massive Regularization2025.07 | 5.14 | 43.56 | |
| CoIBAModel=DeiT3-B-16/224, Training Scheme=DeiT32025.07 | 4.36 | 43.87 |