Legal Judgment Prediction on CMDL
65.5Public Safety AccuracyLegalGraphRAG
Evaluation results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| LegalGraphRAGSize=8B-Inst2026.05 | 65.5 | 66.5 | 59.8 | 65.1 | 58.5 | 63.7 | 70.1 | 72.7 | |
| Legal ∆Size=7B-Inst2026.05 | 58.3 | 61.5 | 51.8 | 55.8 | 50.2 | 54.8 | 65.8 | 64.4 | |
| ADAPTSize=7B-Inst2026.05 | 54.5 | 58.8 | 57.1 | 59.4 | 40.9 | 43.4 | 61.5 | 62.1 | |
| RAPTORSize=8B-Inst2026.05 | 53.8 | 62.6 | 53.6 | 60.1 | 52.5 | 62.8 | 52.1 | 66.9 | |
| HippoRAG2Size=8B-Inst2026.05 | 53.5 | 56.5 | 50.6 | 52.7 | 53.5 | 55 | 62.4 | 62.8 | |
| DISC-LawLLM-7BSize=7B-Inst2026.05 | 49.7 | 53.6 | 39.6 | 52.1 | 30.3 | 49.5 | 48.4 | 63.3 | |
| Naive RAGSize=8B-Inst2026.05 | 45.8 | 57.3 | 44.8 | 55.2 | 46.8 | 58.5 | 49.6 | 57.8 | |
| Qwen-3-8BSize=8B-Inst2026.05 | 44 | 52.3 | 44.7 | 53.1 | 42.7 | 51.9 | 53 | 57.7 | |
| LightRAGSize=8B-Inst2026.05 | 37.9 | 50.1 | 43.2 | 45.1 | 44.2 | 51.3 | 43.7 | 46.9 | |
| G-retrieverSize=8B-Inst2026.05 | 36.8 | 40 | 42.5 | 48.8 | 45.3 | 50.7 | 46.2 | 52.4 | |
| DeepSeek-V3.1Size=∼200B2026.05 | 35 | 64 | 54.7 | 62.7 | 58.2 | 61.9 | 62.5 | 71.6 | |
| Qwen-2.5-7B-InstructSize=7B-Inst2026.05 | 25.8 | 32.4 | 28.7 | 35.8 | 27.2 | 42.1 | 32.8 | 49.6 | |
| Internlm3-8b-instructSize=8B-Inst2026.05 | 25.4 | 32.1 | 35.7 | 37 | 27.5 | 36.2 | 34.1 | 53.6 | |
| Glm-4-9b-chatSize=9B-Inst2026.05 | 23.5 | 34.2 | 23.6 | 40.8 | 19.1 | 37 | 41.5 | 47 | |
| GPT-4o-miniSize=∼8B2026.05 | 18 | 28 | 22.7 | 31.9 | 21.9 | 32 | 35.9 | 50.3 |