Policy Evaluation on PolicyBench Level 3 CN
80.34AccuracyQwQ 32B
Evaluation Results
| Method | Links | |
|---|---|---|
| QwQ 32B2026.04 | 80.34 | |
| o4-mini2026.04 | 79.49 | |
| Gemini 2.5Model Variant=Gemini-2.5-Flash2026.04 | 76.18 | |
| Claude 3.7Model Variant=Claude-3.7-Sonnet2026.04 | 73.82 | |
| Gemini 2.0Model Variant=Gemini-2.0-Flash2026.04 | 73.8 | |
| Deepseek R12026.04 | 73.78 | |
| Claude 3.5Model Variant=Claude-3.5-Sonnet2026.04 | 72.83 | |
| Deepseek V32026.04 | 72.33 | |
| Gemma 3-27B2026.04 | 71.51 | |
| GPT-4o2026.04 | 70.24 | |
| LLaMA 42026.04 | 68.54 |