Selective Classification on ImageNet-Rendition
0.0581AURCEnergy
Evaluation results
| Method | Links | ||
|---|---|---|---|
| EnergyBackbone=DFN CLIP, Evaluation=Zero-shot2025.05 | 0.0581 | 1.06 | |
| SIRCBackbone=DFN CLIP, Evaluation=Zero-shot2025.05 | 0.039 | 0.701 | |
| MDSBackbone=DFN CLIP, Evaluation=Zero-shot2025.05 | 0.0341 | 0.608 | |
| KNNBackbone=DFN CLIP, Evaluation=Zero-shot2025.05 | 0.0251 | 0.439 | |
| MaxLogitBackbone=DFN CLIP, Evaluation=Zero-shot2025.05 | 0.0221 | 0.385 | |
| Δ-MDSBackbone=DFN CLIP, Evaluation=Zero-shot2025.05 | 0.0219 | 0.38 | |
| MSPBackbone=DFN CLIP, Evaluation=Zero-shot2025.05 | 0.02 | 0.344 | |
| Δ-KNNBackbone=DFN CLIP, Evaluation=Zero-shot2025.05 | 0.0142 | 0.237 | |
| Δ-MDS-RLogBackbone=DFN CLIP, Evaluation=Zero-shot2025.05 | 0.0109 | 0.175 | |
| RLogBackbone=DFN CLIP, Evaluation=Zero-shot2025.05 | 0.0081 | 0.122 | |
| Δ-KNN-RLogBackbone=DFN CLIP, Evaluation=Zero-shot2025.05 | 0.0077 | 0.115 |