Evidence Retrieval on Claim-evidence sets (test)
54.19MAPQwen3-Reranker-8B
Evaluation results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Qwen3-Reranker-8BEvaluation Setting=Fine-tuned2026.01 | 54.19 | 83.44 | 14.84 | 27.32 | 48.26 | 85.19 | 83.33 | 82.11 | |
| E2Rank-0.6BEvaluation Setting=Fine-tuned2026.01 | 52.37 | 86.57 | 14.69 | 25.96 | 46.34 | 86.75 | 83.95 | 82.06 | |
| E2Rank-0.6BEvaluation Setting=Zero-shot2026.01 | 47.57 | 85.54 | 12.86 | 22.8 | 41.18 | 85.19 | 81.75 | 79.2 | |
| E2Rank-8BEvaluation Setting=Zero-shot2026.01 | 47.54 | 85.77 | 12.57 | 22.78 | 41.36 | 85.36 | 81.8 | 79.25 | |
| E2Rank-4BEvaluation Setting=Zero-shot2026.01 | 47.41 | 86.06 | 12.67 | 22.68 | 41.07 | 85.16 | 81.63 | 79.13 | |
| Qwen3-Reranker-8BEvaluation Setting=Zero-shot2026.01 | 45.88 | 71.72 | 11.09 | 21.52 | 41.49 | 75.34 | 74.65 | 74.28 | |
| MiniLM-L6-v2Evaluation Setting=Fine-tuned2026.01 | 45.51 | 74.62 | 11.33 | 21.37 | 40 | 77.41 | 76.14 | 75.02 | |
| Qwen3-Reranker-4BEvaluation Setting=Zero-shot2026.01 | 44.92 | 68.74 | 10.93 | 21.02 | 40.71 | 73.37 | 73.13 | 73.12 | |
| Qwen3-Reranker-0.6BEvaluation Setting=Zero-shot2026.01 | 44.19 | 66.73 | 10.54 | 20.71 | 39.77 | 72.07 | 72.05 | 72.14 | |
| MiniLM-L6-v2Evaluation Setting=Zero-shot2026.01 | 44.08 | 66.12 | 10.37 | 20.46 | 39.72 | 71.48 | 71.66 | 71.82 | |
| bge-reranker-v2-m3Evaluation Setting=Zero-shot2026.01 | 43.99 | 65.3 | 10.35 | 20.23 | 39.73 | 70.94 | 71.25 | 71.55 | |
| gte-reranker-baseEvaluation Setting=Zero-shot2026.01 | 43.99 | 65.8 | 10.47 | 20.4 | 39.43 | 71.54 | 71.62 | 71.84 | |
| GritLM-7BEvaluation Setting=Zero-shot2026.01 | 43.15 | 66.17 | 10.16 | 19.47 | 38.22 | 70.83 | 71.39 | 71.36 |