Robustness Evaluation on Perturbation Dataset
62.56Change AccuracyL4L
Evaluation Results
| Method | Links | |
|---|---|---|
| L4L2025.11 | 62.56 | |
| GPT-5.22025.11 | 59.63 | |
| DeepSeek v32025.11 | 55.93 | |
| Claude 4 Sonnet2025.11 | 51.5 | |
| GPT-4o2025.11 | 50.67 | |
| GPT o4-mini2025.11 | 46.33 | |
| DISC-LawLLM2025.11 | 23.17 | |
| LexiLaw2025.11 | 0 |