OOD Detection on ImageNet V2 (test)
60.78AUROCMM++
Evaluation results
| Method | Links | ||
|---|---|---|---|
| MM++Backbone=ViT-B/16, ID Dataset=ImageNet-LT2026.06 | 60.78 | 88.94 | |
| Maha++Backbone=ViT-B/16, ID Dataset=ImageNet-LT2026.06 | 59.52 | 89.37 | |
| rMahaBackbone=ViT-B/16, ID Dataset=ImageNet-LT2026.06 | 59.18 | 89.6 | |
| rMaha++Backbone=ViT-B/16, ID Dataset=ImageNet-LT2026.06 | 59.08 | 89.34 | |
| MSPBackbone=ViT-B/16, ID Dataset=ImageNet-LT2026.06 | 58.73 | 90.14 | |
| ReActBackbone=ViT-B/16, ID Dataset=ImageNet-LT2026.06 | 58.65 | 89.91 | |
| X-Maha†Backbone=ViT-B/16, ID Dataset=ImageNet-LT2026.06 | 58.36 | 91.36 | |
| MahaBackbone=ViT-B/16, ID Dataset=ImageNet-LT2026.06 | 58.29 | 89.76 | |
| EnergyBackbone=ViT-B/16, ID Dataset=ImageNet-LT2026.06 | 57.29 | 90.85 | |
| ODINBackbone=ViT-B/16, ID Dataset=ImageNet-LT2026.06 | 54.82 | 92.1 | |
| KNNBackbone=ViT-B/16, ID Dataset=ImageNet-LT2026.06 | 49.8 | 94.34 |