Evidence Retrieval on Evidence Inference Dataset 2.0 (local)
20.16Recall@1CERA
Evaluation Results
| Method | Links | ||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| CERA2026.05 | 20.16 | 41.69 | 56.08 | 74.42 | 89.47 | 98.02 | 25.93 | 18.63 | 15.04 | 10.1 | 6.07 | 2.66 | 34.96 | 41.29 | 47.76 | 51.91 | 53.79 | 25.93 | 30.57 | 34.37 | 37.45 | 38.83 | 39.21 | 42.96 | |
| HardNegbaseselection_strategy=hard-negative2026.05 | 12.48 | 30.96 | 43.01 | 63.56 | 83.6 | 96.27 | 16.59 | 13.55 | 11.42 | 8.53 | 5.64 | 2.61 | 24.61 | 29.95 | 37.17 | 42.67 | 45.47 | 16.59 | 20.7 | 23.81 | 27.12 | 28.88 | 29.46 | 32.88 | |
| Contriever2026.05 | 2.14 | 6.4 | 11.69 | 20.35 | 33.92 | 60.27 | 3.02 | 2.99 | 3.15 | 2.77 | 2.31 | 1.64 | 4.93 | 7.22 | 10.27 | 13.95 | 19.56 | 3.02 | 3.87 | 5.09 | 6.32 | 7.35 | 8.31 | 10.63 |