Open-ended QA Response Ranking on AlignBench Minos
47.68K ScoreMinosEval
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| MinosEvalManual Factoid Classification=true2025.06 | 47.68 | 57.38 | 57.09 | 86.62 | |
| MinosEval2025.06 | 45.28 | 54.89 | 56.3 | 86.28 | |
| Listwise2025.06 | 41.47 | 51.46 | 54.53 | 85.64 | |
| Pairwise2025.06 | 38.66 | 47.89 | 51.59 | 84.85 | |
| LINKAGE2025.06 | 35.75 | 43.82 | 52.97 | 84.88 | |
| Pointwise2025.06 | 32.94 | 41.37 | 47.41 | 83.59 | |
| BLEU2025.06 | 15.41 | 19.47 | 37.81 | 79.98 | |
| BERTScore2025.06 | 13.62 | 18.12 | 37.93 | 80.09 | |
| ROUGE-L2025.06 | 13.38 | 16.69 | 36.24 | 79.93 |