Safety Knowledge Assessment on Risk Knowledge
91.9Overall ScoreQwen3.5-Plus
Evaluation Results
| Method | Links | |
|---|---|---|
| Qwen3.5-PlusModel Size=397B2026.06 | 91.9 | |
| Yuvion VL-32BProtocol=Reasoning2026.06 | 91.6 | |
| GPT-5.4Model Size=≈1.5T2026.06 | 91.1 | |
| Yuvion VL-8BProtocol=Reasoning2026.06 | 91 | |
| Yuvion VL-32BProtocol=Instruct2026.06 | 90.6 | |
| Qwen3-VL-32BProtocol=Thinking2026.06 | 90.5 | |
| Qwen3-VL-8BProtocol=Thinking2026.06 | 90.2 | |
| Yuvion VL-8BProtocol=Instruct2026.06 | 89.8 | |
| K2.5Model Size=≈1T2026.06 | 89.3 | |
| Qwen3-VL-32BProtocol=Instruct2026.06 | 88.2 | |
| Qwen3-VL-8BProtocol=Instruct2026.06 | 86.6 | |
| GLM-5Model Size=≈700B2026.06 | 86.1 |