List your company

Safety and Helpfulness Evaluation on Beavertails

92.03Safety ScoreSAFER-STEER
26.634843.612460.5977.5676Mar 18, 2026

Evaluation results

MethodLinks
SAFER-STEERBase Model=Qwen2.5VL-7B, Method Category=Dynamic Methods2026.0392.0386.83
SAFER-STEERBase Model=Qwen2.5VL-3B, Method Category=Dynamic Methods2026.0388.8985.97
SFTBase Model=Qwen2.5VL-7B, Method Category=Dynamic Methods2026.0388.8186.1
SFTBase Model=Qwen2.5VL-3B, Method Category=Dynamic Methods2026.0386.9582.88
TISBase Model=Qwen2.5VL-7B, Method Category=Static Methods2026.0385.5977.8
SPA-VLBase Model=Qwen2.5VL-7B, Method Category=Static Methods2026.0380.8577.76
SPA-VLBase Model=Qwen2.5VL-3B, Method Category=Static Methods2026.0380.8176.36
TISBase Model=Qwen2.5VL-3B, Method Category=Static Methods2026.0379.1573.73
SaFeR-VLMBase Model=Qwen2.5VL-7B, Method Category=Static Methods2026.0376.7875.93
SaFeR-VLMBase Model=Qwen2.5VL-3B, Method Category=Static Methods2026.0376.4481.19
MMDPOBase Model=Qwen2.5VL-7B, Method Category=Static Methods2026.0368.5473.03
Qwen2.5VL-7BBase Model=Qwen2.5VL-7B, Method Category=Base2026.0355.9358.98
GRLHF-VBase Model=Qwen2.5VL-7B, Method Category=Static Methods2026.0349.4955.59
MMDPOBase Model=Qwen2.5VL-3B, Method Category=Static Methods2026.0347.1854.19
SafeRLHF-VBase Model=Qwen2.5VL-7B, Method Category=Static Methods2026.0346.138.81
Qwen2.5VL-3BBase Model=Qwen2.5VL-3B, Method Category=Base2026.0335.3141.93
GRLHF-VBase Model=Qwen2.5VL-3B, Method Category=Static Methods2026.0332.5436.95
SafeRLHF-VBase Model=Qwen2.5VL-3B, Method Category=Static Methods2026.0329.1524.07