Critique and Refinement Evaluation Overall
84.3CU ScoreRCO
Evaluation results
| Method | Links | ||
|---|---|---|---|
| RCOBackbone=LLaMA-3-8B-Instruct2025.06 | 84.3 | 6.62 | |
| Base modelBackbone=LLaMA-3-70B-Instruct2025.06 | 82.2 | 6.44 | |
| RCOBackbone=LLaMA-2-13B-Chat2025.06 | 81.9 | 6.42 | |
| RCOBackbone=LLaMA-2-7B-Chat2025.06 | 81.1 | 6.49 | |
| RCOBackbone=Auto-J-13B2025.06 | 78.8 | 6.36 | |
| RCOBackbone=UltraCM-13B2025.06 | 78.8 | 6.29 | |
| Base modelBackbone=LLaMA-3-8B-Instruct2025.06 | 75.8 | 6.33 | |
| DPCOBackbone=Auto-J-13B2025.06 | 74.2 | 6.16 | |
| DPCOBackbone=UltraCM-13B2025.06 | 73.8 | 6.06 | |
| Base modelBackbone=LLaMA-2-13B-Chat2025.06 | 73.5 | 6.17 | |
| DPCOBackbone=LLaMA-2-13B-Chat2025.06 | 73.5 | 6.08 | |
| DPCOBackbone=LLaMA-3-8B-Instruct2025.06 | 72.9 | 6.07 | |
| Self-refinement2025.06 | 72.6 | 5.97 | |
| Base modelBackbone=LLaMA-2-70B-Chat2025.06 | 72.4 | 6.21 | |
| DPCOBackbone=LLaMA-2-7B-Chat2025.06 | 72.3 | 6.16 | |
| Base modelBackbone=LLaMA-2-7B-Chat2025.06 | 70.8 | 6.08 | |
| Base modelBackbone=Auto-J-13B2025.06 | 69.7 | 6 | |
| Base modelBackbone=UltraCM-13B2025.06 | 63.6 | 5.83 | |
| Initial Answer2025.06 | 46.9 | 5.25 | |
| Aligner2025.06 | 46.9 | 5.4 |