Policy Question Answering on PolicyBench Chinese
65.33AccuracyQwQ-32B
Evaluation Results
| Method | Links | |
|---|---|---|
| QwQ-32B2026.04 | 65.33 | |
| Gemini-2.5-Flash2026.04 | 63.6 | |
| Claude-3.7-sonnet2026.04 | 63.19 | |
| Deepseek-R12026.04 | 62.24 | |
| Claude-3.5-Sonnet2026.04 | 62.11 | |
| o4-mini2026.04 | 60.41 | |
| Gemini-2.0-Flash2026.04 | 59.35 | |
| Deepseek-V32026.04 | 58.82 | |
| LLaMA-42026.04 | 58.3 | |
| GPT-4o2026.04 | 57.53 | |
| Gemma 3-27B2026.04 | 56.27 |