Legal Contract Revision on MAUD
86.89CQ ScoreRCBSF
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| RCBSFBackbone=Qwen2.5-7B-Chat2026.04 | 86.89 | 90.83 | 87.35 | |
| RCBSFBackbone=Qwen2.5-7B-Base2026.04 | 83.45 | 87.12 | 84.23 | |
| RCBSFBackbone=LawLLM-7B2026.04 | 80.77 | 76.34 | 79.56 | |
| IterationBackbone=Qwen2.5-7B-Chat2026.04 | 80.45 | 84.56 | 83.23 | |
| RAGBackbone=Qwen2.5-7B-Chat2026.04 | 79.56 | 82.11 | 81.45 | |
| RCBSFBackbone=Mistral-7B2026.04 | 78.91 | 75.23 | 78.55 | |
| IterationBackbone=Qwen2.5-7B-Base2026.04 | 77.45 | 80.89 | 79.56 | |
| RAGBackbone=Qwen2.5-7B-Base2026.04 | 76.56 | 78.45 | 78.12 | |
| RAGBackbone=LawLLM-7B2026.04 | 75.34 | 69.89 | 76.23 | |
| CoTBackbone=Qwen2.5-7B-Chat2026.04 | 75.12 | 78.34 | 76.45 | |
| RCBSFBackbone=LexiLaw-6B2026.04 | 74.88 | 70.45 | 73.78 | |
| IterationBackbone=LawLLM-7B2026.04 | 74.23 | 72.45 | 75.67 | |
| IterationBackbone=Mistral-7B2026.04 | 73.45 | 70.89 | 74.56 | |
| StandardBackbone=Qwen2.5-7B-Chat2026.04 | 73.45 | 76.12 | 74.23 | |
| CoTBackbone=Qwen2.5-7B-Base2026.04 | 72.12 | 74.56 | 73.89 | |
| RAGBackbone=Mistral-7B2026.04 | 71.56 | 67.23 | 71.88 | |
| IterationBackbone=LexiLaw-6B2026.04 | 70.23 | 67.56 | 67.89 | |
| StandardBackbone=Qwen2.5-7B-Base2026.04 | 69.88 | 72.34 | 71.45 | |
| CoTBackbone=LawLLM-7B2026.04 | 69.88 | 65.12 | 67.45 | |
| RAGBackbone=LexiLaw-6B2026.04 | 68.89 | 66.78 | 69.56 | |
| CoTBackbone=Mistral-7B2026.04 | 67.89 | 63.45 | 66.23 | |
| StandardBackbone=LawLLM-7B2026.04 | 67.45 | 62.56 | 66.89 | |
| CoTBackbone=LexiLaw-6B2026.04 | 66.45 | 65.89 | 66.23 | |
| StandardBackbone=Mistral-7B2026.04 | 65.34 | 60.88 | 64.12 | |
| StandardBackbone=LexiLaw-6B2026.04 | 65.15 | 65.11 | 65.21 |