Unsupervised OOD detection on The Pile ID Long-COVID OOD (test)
91.79AUROCAP-OOD
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| AP-OODBase Model=Pythia-160M, ID Dataset=The Pile, OOD Type=Input OOD2026.02 | 91.79 | 40.34 | |
| PerplexityBase Model=Pythia-160M, ID Dataset=The Pile, OOD Type=Input OOD2026.02 | 89.51 | 68.6 | |
| MahalanobisBase Model=Pythia-160M, ID Dataset=The Pile, OOD Type=Input OOD2026.02 | 75.5 | 98.07 | |
| Deep SVDDBase Model=Pythia-160M, ID Dataset=The Pile, OOD Type=Input OOD2026.02 | 72.54 | 99.52 | |
| KNNBase Model=Pythia-160M, ID Dataset=The Pile, OOD Type=Input OOD2026.02 | 70.62 | 99.03 |