OOD Detection on Law (OOD) vs WMT15 En-Fr (ID)
63.38AUROCDeep SAD
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Deep SADOOD level=Output, Learning paradigm=Supervised, Backbone=Transformer (base)2026.02 | 63.38 | 91.53 | |
| AP-OODOOD level=Input, Learning paradigm=Supervised, Backbone=Transformer (base)2026.02 | 60.3 | 95.91 | |
| AP-OODOOD level=Output, Learning paradigm=Supervised, Backbone=Transformer (base)2026.02 | 56.6 | 97.18 | |
| Binary logitsOOD level=Output, Learning paradigm=Supervised, Backbone=Transformer (base)2026.02 | 54.28 | 96.42 | |
| Deep SADOOD level=Input, Learning paradigm=Supervised, Backbone=Transformer (base)2026.02 | 52.11 | 95.95 | |
| Binary logitsOOD level=Input, Learning paradigm=Supervised, Backbone=Transformer (base)2026.02 | 51.23 | 95.96 | |
| Relative MahalanobisOOD level=Output, Learning paradigm=Supervised, Backbone=Transformer (base)2026.02 | 30.41 | 96.72 | |
| Relative MahalanobisOOD level=Input, Learning paradigm=Supervised, Backbone=Transformer (base)2026.02 | 30.11 | 96.99 |