List your company

Multimodal Safety Evaluation on ToolkitBench

2.49Safety ScoreQwen2.5-VL-7B + SFT+DPO+GRPO
1.53321.78162.032.2784Mar 3, 2026

Evaluation results

MethodLinks
Qwen2.5-VL-7B + SFT+DPO+GRPOBase Model=Qwen2.5-VL-7B, Stage=SFT+DPO+GRPO2026.032.492.52.45
Qwen2.5-VL-7B + SFT+GRPOBase Model=Qwen2.5-VL-7B, Stage=SFT+GRPO2026.032.472.352.31
Qwen2.5-VL-3B + SFT+DPO+GRPOBase Model=Qwen2.5-VL-3B, Stage=SFT+DPO+GRPO2026.032.452.212.23
Qwen2.5-VL-3B + SFT+GRPOBase Model=Qwen2.5-VL-3B, Stage=SFT+GRPO2026.032.392.122.15
Qwen2.5-VL-7B + TISBase Model=Qwen2.5-VL-7B, Enhancement=TIS2026.032.371.941.98
Qwen2.5-VL-7B + SFT+DPOBase Model=Qwen2.5-VL-7B, Stage=SFT+DPO2026.032.292.342.22
Qwen2.5-VL-7B + SaFeR-VLMBase Model=Qwen2.5-VL-7B, Method=SaFeR-VLM2026.032.261.621.66
Qwen2.5-VL-7B + SPA-VLBase Model=Qwen2.5-VL-7B, Enhancement=SPA-VL2026.032.231.971.45
Qwen2.5-VL-3B + SFT+DPOBase Model=Qwen2.5-VL-3B, Stage=SFT+DPO2026.032.181.941.85
Qwen2.5-VL-7B + SFTBase Model=Qwen2.5-VL-7B, Stage=SFT2026.032.152.152.01
Qwen2.5-VL-3B + SPA-VLBase Model=Qwen2.5-VL-3B, Enhancement=SPA-VL2026.032.131.781.27
Qwen2.5-VL-7B + SIABase Model=Qwen2.5-VL-7B, Enhancement=SIA2026.032.131.791.44
Qwen2.5-VL-3B + SaFeR-VLMBase Model=Qwen2.5-VL-3B, Method=SaFeR-VLM2026.032.081.461.43
Qwen2.5-VL-3B + SFTBase Model=Qwen2.5-VL-3B, Stage=SFT2026.032.071.871.74
Qwen2.5-VL-7BBase Model=Qwen2.5-VL-7B2026.032.072.051.54
Qwen2.5-VL-7B + ECSOBase Model=Qwen2.5-VL-7B, Enhancement=ECSO2026.032.042.021.52
Qwen2.5-VL-3B + TISBase Model=Qwen2.5-VL-3B, Enhancement=TIS2026.031.971.451.61
Qwen2.5-VL-7B + VLGuardBase Model=Qwen2.5-VL-7B, Enhancement=VLGuard2026.031.771.191.11
Qwen2.5-VL-3B + VLGuardBase Model=Qwen2.5-VL-3B, Enhancement=VLGuard2026.031.691.11.09
Qwen2.5-VL-3B + SIABase Model=Qwen2.5-VL-3B, Enhancement=SIA2026.031.621.461.01
Qwen2.5-VL-3B + ECSOBase Model=Qwen2.5-VL-3B, Enhancement=ECSO2026.031.581.661.13
Qwen2.5-VL-3BBase Model=Qwen2.5-VL-3B2026.031.571.741.18