Safety Evaluation on SIUO
77.4Safe RateSafework-R1-7B
Evaluation Results
| Method | Links | |
|---|---|---|
| Safework-R1-7BModel=Safework-R1-7B2025.12 | 77.4 | |
| Gemini-2.5-proModel=Gemini-2.5-pro2025.12 | 76.7 | |
| TRRBackbone=Qwen2.5VL-7B, Alignment Method=TRR2025.12 | 76.2 | |
| MSR-AlignBackbone=Qwen2.5VL-32B, Alignment Method=MSR-Align2025.12 | 73.2 | |
| TRRBackbone=Qwen2.5VL-32B, Alignment Method=TRR2025.12 | 71.5 | |
| MSR-AlignBackbone=Qwen2.5VL-7B, Alignment Method=MSR-Align2025.12 | 70.7 | |
| TiSBackbone=Qwen2.5VL-32B, Alignment Method=TiS2025.12 | 67.1 | |
| Claude-3.5-SonnetModel=Claude-3.5-Sonnet2025.12 | 56.7 | |
| GPT-4oModel=GPT-4o2025.12 | 51.8 | |
| Qwen2.5VL-32BBackbone=Qwen2.5VL-32B, Alignment Method=None2025.12 | 42.7 | |
| TiSBackbone=Qwen2.5VL-7B, Alignment Method=TiS2025.12 | 37.8 | |
| Qwen2.5VL-7BBackbone=Qwen2.5VL-7B, Alignment Method=None2025.12 | 30.8 | |
| TGAModel Size=7B2024.10 | 30.77 | |
| LLaVA-v1.5Model Size=13B2024.10 | 22.16 | |
| LLaVA-v1.5Model Size=7B2024.10 | 21.56 |