Critique Quality Evaluation on QA
75Win RateRCO
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| RCOComparison Target=LLaMa-3-70B-Instruct2025.06 | 75 | 17.5 | 7.5 | 67.5 | — | |
| RCOComparison Target=LLaMa-3-70B-Instruct2025.06 | 62.5 | 15 | 22.5 | — | 40 | |
| RCOComparison Target=LLaMa-3-8B-Instruct + MSE + DPO2025.06 | 47.5 | 15 | 37.5 | — | 10 | |
| RCOComparison Target=LLaMa-3-8B-Instruct2025.06 | 42.5 | 35 | 22.5 | 20 | — | |
| RCOComparison Target=LLaMa-3-8B-Instruct2025.06 | 37.5 | 20 | 42.5 | — | -5 | |
| RCOComparison Target=LLaMa-3-8B-Instruct + MSE + DPO2025.06 | 32.5 | 27.5 | 40 | -7.5 | — |