Input OOD Detection on Newsroom
99.89AUROCAP-OOD
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| AP-OODBackbone=PEGASUS_LARGE, ID Dataset=XSUM2026.02 | 99.89 | 0.39 | |
| Binary logitsBackbone=PEGASUS_LARGE, ID Dataset=XSUM2026.02 | 99.47 | 2.04 | |
| Deep SADBackbone=PEGASUS_LARGE, ID Dataset=XSUM2026.02 | 99.44 | 1.49 | |
| AP-OODBackbone=PEGASUS-LARGE, ID Dataset=XSUM, Evaluation Protocol=Input OOD2026.02 | 98.54 | 5.78 | |
| Relative MahalanobisBackbone=PEGASUS_LARGE, ID Dataset=XSUM2026.02 | 91.71 | 28.41 | |
| Deep SVDDBackbone=PEGASUS-LARGE, ID Dataset=XSUM, Evaluation Protocol=Input OOD2026.02 | 91.36 | 36.05 | |
| MahalanobisBackbone=PEGASUS-LARGE, ID Dataset=XSUM, Evaluation Protocol=Input OOD2026.02 | 86.4 | 64.02 | |
| KNNBackbone=PEGASUS-LARGE, ID Dataset=XSUM, Evaluation Protocol=Input OOD2026.02 | 73.83 | 88.49 |