List your company

Multi-turn MLLM Safety Evaluation on STEER-DyS Multi-turn

61.36Safety ScoreSAFER-STEER
3.005618.155333.30548.4547Mar 18, 2026

Evaluation results

MethodLinks
SAFER-STEERBackbone=Qwen2.5VL-7B, Alignment Strategy=SAFER-STEER2026.0361.3655.42
SFTBackbone=Qwen2.5VL-7B, Alignment Strategy=SFT2026.0350.4651.68
SAFER-STEERBackbone=Qwen2.5VL-3B, Alignment Strategy=SAFER-STEER2026.0349.455.4
SaFeR-VLMBackbone=Qwen2.5VL-7B, Alignment Strategy=SaFeR-VLM2026.0349.3847.85
SFTBackbone=Qwen2.5VL-3B, Alignment Strategy=SFT2026.0335.1738.84
SaFeR-VLMBackbone=Qwen2.5VL-3B, Alignment Strategy=SaFeR-VLM2026.0332.4235.17
TISBackbone=Qwen2.5VL-7B, Alignment Strategy=TIS2026.0332.1621.29
SPA-VLBackbone=Qwen2.5VL-7B, Alignment Strategy=SPA-VL2026.0329.4735.58
MMDPOBackbone=Qwen2.5VL-7B, Alignment Strategy=MMDPO2026.0326.6444.47
TISBackbone=Qwen2.5VL-3B, Alignment Strategy=TIS2026.0324.9216.82
Qwen2.5VL-7BBackbone=Qwen2.5VL-7B, Alignment Strategy=Base2026.0322.2738.7
GRLHF-VBackbone=Qwen2.5VL-7B, Alignment Strategy=GRLHF-V2026.0318.8734.66
SPA-VLBackbone=Qwen2.5VL-3B, Alignment Strategy=SPA-VL2026.0316.7419.69
SafeRLHF-VBackbone=Qwen2.5VL-7B, Alignment Strategy=SafeRLHF-V2026.0313.4312.5
MMDPOBackbone=Qwen2.5VL-3B, Alignment Strategy=MMDPO2026.0311.3318.22
GRLHF-VBackbone=Qwen2.5VL-3B, Alignment Strategy=GRLHF-V2026.039.6320.64
Qwen2.5VL-3BBackbone=Qwen2.5VL-3B, Alignment Strategy=Base2026.039.419.44
SafeRLHF-VBackbone=Qwen2.5VL-3B, Alignment Strategy=SafeRLHF-V2026.035.256.33