Legal Contract Revision on ALL Avg
86.87CQ ScoreRCBSF
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| RCBSFBackbone=Qwen2.5-7B-Chat2026.04 | 86.87 | 84.21 | 87.29 | |
| RCBSFBackbone=Qwen2.5-7B-Base2026.04 | 83.67 | 80.71 | 84.23 | |
| RCBSFBackbone=LawLLM-7B2026.04 | 82.64 | 78.29 | 82.54 | |
| RCBSFBackbone=Mistral-7B2026.04 | 80.23 | 76.59 | 80.57 | |
| RAGBackbone=Qwen2.5-7B-Chat2026.04 | 79.81 | 77.25 | 81.73 | |
| IterationBackbone=Qwen2.5-7B-Chat2026.04 | 79.76 | 79.56 | 83.4 | |
| RCBSFBackbone=LexiLaw-6B2026.04 | 77.3 | 71.59 | 76.67 | |
| IterationBackbone=LawLLM-7B2026.04 | 77.01 | 75.23 | 79.36 | |
| RAGBackbone=Qwen2.5-7B-Base2026.04 | 76.84 | 74.01 | 78.65 | |
| RAGBackbone=LawLLM-7B2026.04 | 76.81 | 72.25 | 77.73 | |
| IterationBackbone=Qwen2.5-7B-Base2026.04 | 76.78 | 76.09 | 80.17 | |
| CoTBackbone=Qwen2.5-7B-Chat2026.04 | 75.67 | 73.26 | 76.28 | |
| IterationBackbone=Mistral-7B2026.04 | 75.56 | 72.51 | 76.31 | |
| StandardBackbone=Qwen2.5-7B-Chat2026.04 | 73.45 | 70.81 | 74.31 | |
| RAGBackbone=Mistral-7B2026.04 | 73.23 | 69.04 | 73.69 | |
| CoTBackbone=Qwen2.5-7B-Base2026.04 | 72.76 | 70.09 | 73.45 | |
| RAGBackbone=LexiLaw-6B2026.04 | 71.67 | 67.67 | 72.09 | |
| CoTBackbone=LawLLM-7B2026.04 | 71.03 | 66.9 | 68.65 | |
| StandardBackbone=Qwen2.5-7B-Base2026.04 | 70.42 | 67.7 | 71.26 | |
| IterationBackbone=LexiLaw-6B2026.04 | 70.31 | 68.72 | 69.92 | |
| CoTBackbone=Mistral-7B2026.04 | 69.17 | 65.11 | 67.56 | |
| StandardBackbone=LawLLM-7B2026.04 | 68.67 | 64.03 | 69.2 | |
| CoTBackbone=LexiLaw-6B2026.04 | 67.25 | 66.39 | 66.89 | |
| StandardBackbone=Mistral-7B2026.04 | 66.92 | 61.75 | 65.92 | |
| StandardBackbone=LexiLaw-6B2026.04 | 65.4 | 65.13 | 65.5 |