Policy Evaluation on PolicyBench Level 3 US
77Accuracyo4-mini
Evaluation Results
| Method | Links | |
|---|---|---|
| o4-mini2026.04 | 77 | |
| Deepseek R12026.04 | 74.6 | |
| QwQ 32B2026.04 | 69.9 | |
| Gemini 2.5Model Variant=Gemini-2.5-Flash2026.04 | 69.44 | |
| Deepseek V32026.04 | 69.39 | |
| Claude 3.5Model Variant=Claude-3.5-Sonnet2026.04 | 68.47 | |
| Gemma 3-27B2026.04 | 68.37 | |
| Claude 3.7Model Variant=Claude-3.7-Sonnet2026.04 | 68.28 | |
| GPT-4o2026.04 | 68.13 | |
| Gemini 2.0Model Variant=Gemini-2.0-Flash2026.04 | 66.55 | |
| LLaMA 42026.04 | 66.41 |