Reranking on PD Dataset (All)
77.9PrecisionZeroshot Llama 3.1
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Zeroshot Llama 3.1Evaluation Protocol=Zero-shot2026.01 | 77.9 | 81.7 | 79.7 | 72.7 | |
| bge-reranker-largeEvaluation Protocol=Zero-shot2026.01 | 77.5 | 56.3 | 65.2 | 60.5 | |
| bge-reranker-baseEvaluation Protocol=Zero-shot2026.01 | 77.4 | 59.8 | 67.5 | 62.1 | |
| jina-reranker-v2-base-multilingualEvaluation Protocol=Zero-shot2026.01 | 76.4 | 65.1 | 70.3 | 63.8 | |
| bge-reranker-v2-m3Evaluation Protocol=Zero-shot2026.01 | 75.7 | 73.6 | 74.6 | 67.1 | |
| Zeroshot Qwen3-Reranker-8BEvaluation Protocol=Zero-shot2026.01 | 73.7 | 98.7 | 84.4 | 76 | |
| majority baselineEvaluation Protocol=Baseline2026.01 | 66.8 | 1 | 80.1 | 66.8 |