Legal Contract Revision on CUAD
85.82CQRCBSF
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| RCBSFBackbone=Qwen2.5-7B-Chat2026.04 | 85.82 | 81.72 | 86.42 | |
| RCBSFBackbone=Qwen2.5-7B-Base2026.04 | 82.67 | 78.34 | 83.45 | |
| RCBSFBackbone=LawLLM-7B2026.04 | 82.56 | 79.23 | 83.12 | |
| RAGBackbone=Qwen2.5-7B-Chat2026.04 | 79.34 | 74.56 | 80.78 | |
| RCBSFBackbone=Mistral-7B2026.04 | 78.92 | 77.11 | 81.56 | |
| IterationBackbone=Qwen2.5-7B-Chat2026.04 | 78.12 | 76.89 | 82.45 | |
| RAGBackbone=LawLLM-7B2026.04 | 77.56 | 73.45 | 78.12 | |
| RAGBackbone=Qwen2.5-7B-Base2026.04 | 76.45 | 71.23 | 77.56 | |
| RCBSFBackbone=LexiLaw-6B2026.04 | 76.22 | 71.45 | 77.34 | |
| IterationBackbone=LawLLM-7B2026.04 | 76.12 | 76.89 | 80.45 | |
| IterationBackbone=Qwen2.5-7B-Base2026.04 | 75.23 | 73.45 | 79.12 | |
| IterationBackbone=Mistral-7B2026.04 | 75.23 | 72.45 | 77.12 | |
| CoTBackbone=Qwen2.5-7B-Chat2026.04 | 75.23 | 71.34 | 75.89 | |
| StandardBackbone=Qwen2.5-7B-Chat2026.04 | 72.89 | 68.56 | 73.45 | |
| CoTBackbone=Qwen2.5-7B-Base2026.04 | 72.56 | 68.34 | 72.89 | |
| RAGBackbone=Mistral-7B2026.04 | 72.45 | 69.34 | 73.56 | |
| RAGBackbone=LexiLaw-6B2026.04 | 72.45 | 67.89 | 72.23 | |
| CoTBackbone=LawLLM-7B2026.04 | 70.45 | 66.89 | 68.23 | |
| StandardBackbone=Qwen2.5-7B-Base2026.04 | 70.23 | 65.67 | 70.12 | |
| IterationBackbone=LexiLaw-6B2026.04 | 70.12 | 69.33 | 70.78 | |
| CoTBackbone=Mistral-7B2026.04 | 68.12 | 65.88 | 67.99 | |
| StandardBackbone=LawLLM-7B2026.04 | 68.12 | 64.33 | 69.12 | |
| CoTBackbone=LexiLaw-6B2026.04 | 67.56 | 66.78 | 66.99 | |
| StandardBackbone=Mistral-7B2026.04 | 66.78 | 62.56 | 66.45 | |
| StandardBackbone=LexiLaw-6B2026.04 | 65.78 | 65.22 | 65.56 |