Output OOD Detection on CNN/DM
99.46AUROCAP-OOD
Evaluation results
| Method | Links | ||
|---|---|---|---|
| AP-OODBackbone=PEGASUS_LARGE, ID Dataset=XSUM2026.02 | 99.46 | 1.96 | |
| Binary logitsBackbone=PEGASUS_LARGE, ID Dataset=XSUM2026.02 | 98.47 | 5.77 | |
| Deep SADBackbone=PEGASUS_LARGE, ID Dataset=XSUM2026.02 | 97.53 | 8.22 | |
| Relative MahalanobisBackbone=PEGASUS_LARGE, ID Dataset=XSUM2026.02 | 93.42 | 25.22 |