OOD Detection on ndt2015 OOD WMT15 En-Fr ID (test)
0.9189AUROCAP-OOD
Evaluation results
| Method | Links | ||
|---|---|---|---|
| AP-OODOOD level=Input, Learning paradigm=Supervised, Backbone=Transformer (base)2026.02 | 0.9189 | 0.4953 | |
| AP-OODOOD level=Output, Learning paradigm=Supervised, Backbone=Transformer (base)2026.02 | 0.9128 | 0.5087 | |
| Deep SADOOD level=Input, Learning paradigm=Supervised, Backbone=Transformer (base)2026.02 | 0.9099 | 0.5733 | |
| Binary logitsOOD level=Input, Learning paradigm=Supervised, Backbone=Transformer (base)2026.02 | 0.905 | 0.5563 | |
| Binary logitsOOD level=Output, Learning paradigm=Supervised, Backbone=Transformer (base)2026.02 | 0.8991 | 0.5699 | |
| Relative MahalanobisOOD level=Output, Learning paradigm=Supervised, Backbone=Transformer (base)2026.02 | 0.877 | 0.6853 | |
| Relative MahalanobisOOD level=Input, Learning paradigm=Supervised, Backbone=Transformer (base)2026.02 | 0.8763 | 0.6783 | |
| AP-OODBackbone=Transformer (base), Training Data=WMT15 En-Fr, Evaluation Side=Output, Unsupervised=true2026.02 | 0.7541 | 0.8156 | |
| PerplexityBackbone=Transformer (base), Training Data=WMT15 En-Fr, Evaluation Side=Output, Unsupervised=true2026.02 | 0.7343 | 0.8567 | |
| EntropyBackbone=Transformer (base), Training Data=WMT15 En-Fr, Evaluation Side=Output, Unsupervised=true2026.02 | 0.7257 | 0.7763 | |
| AP-OODBackbone=Transformer (base), Training Data=WMT15 En-Fr, Evaluation Side=Input, Unsupervised=true2026.02 | 0.695 | 0.8907 | |
| KNNBackbone=Transformer (base), Training Data=WMT15 En-Fr, Evaluation Side=Output, Unsupervised=true2026.02 | 0.6672 | 0.9399 | |
| Deep SADOOD level=Output, Learning paradigm=Supervised, Backbone=Transformer (base)2026.02 | 0.6557 | 0.9307 | |
| KNNBackbone=Transformer (base), Training Data=WMT15 En-Fr, Evaluation Side=Input, Unsupervised=true2026.02 | 0.6516 | 0.9207 | |
| MahalanobisBackbone=Transformer (base), Training Data=WMT15 En-Fr, Evaluation Side=Output, Unsupervised=true2026.02 | 0.6353 | 0.9397 | |
| MahalanobisBackbone=Transformer (base), Training Data=WMT15 En-Fr, Evaluation Side=Input, Unsupervised=true2026.02 | 0.624 | 0.936 | |
| Deep SVDDBackbone=Transformer (base), Training Data=WMT15 En-Fr, Evaluation Side=Output, Unsupervised=true2026.02 | 0.6148 | 0.9391 | |
| Deep SVDDBackbone=Transformer (base), Training Data=WMT15 En-Fr, Evaluation Side=Input, Unsupervised=true2026.02 | 0.6 | 0.9484 |