List your company

Safety Evaluation on MM-SafetyBench (test)

68.95Helpfulness ScoreBeavertails-V help
18.208431.381744.55557.7283Feb 28, 2026

Evaluation results

MethodLinks
Beavertails-V helpBackbone=Llava-1.5-7B2026.0268.9558.2259.14—
Pragma-VLBackbone=Llava-1.5-7B2026.0268.3767.7831.67—
DPOBackbone=Llava-1.5-7B2026.0257.160.6943.4—
Beavertails-V helpBackbone=Qwen2.5-VL-7B2026.0254.9452.3851.07—
Pragma-VLBackbone=Qwen2.5-VL-7B2026.0252.7458.9931.66—
SFTBackbone=Llava-1.5-7B2026.0251.3656.0741.38—
Qwen2.5-VL-7BBackbone=Qwen2.5-VL-7B2026.02505048.75—
Llava-1.5-7BBackbone=Llava-1.5-7B2026.02505056.49—
SPA-VLBackbone=Llava-1.5-7B2026.0247.2654.1744.39—
Beavertails-V allBackbone=Llava-1.5-7B2026.0247.0252.951.53—
SAFE RLHF-vBackbone=Qwen2.5-VL-7B2026.0245.1853.9543.2—
SAFE RLHF-vBackbone=Llava-1.5-7B2026.0244.6953.2748.56—
SFTBackbone=Qwen2.5-VL-7B2026.0243.2953.3639.07—
DPOBackbone=Qwen2.5-VL-7B2026.0239.6651.9736.79—
Beavertails-V allBackbone=Qwen2.5-VL-7B2026.0238.9751.6849.58—
MM-RLHFBackbone=Llava-1.5-7B2026.0237.4443.6946.93—
SPA-VLBackbone=Qwen2.5-VL-7B2026.0235.4349.2648.24—
Beavertails-V harmBackbone=Qwen2.5-VL-7B2026.0227.4445.1843.29—
Beavertails-V harmBackbone=Llava-1.5-7B2026.0225.9538.840.77—
MM-RLHFBackbone=Qwen2.5-VL-7B2026.0220.1632.9235.62—
BaseBackbone=Qwen2.5-Omni-3B2026.05———76.8
BaseBackbone=MiniCPM-o-4.52026.05———69.8
BaseBackbone=Qwen3-Omni-30B-A3B-Instruct2026.05———58.1
ReGapBackbone=Qwen2.5-Omni-3B2026.05———14.2
ReGapBackbone=MiniCPM-o-4.52026.05———65.4
ReGapBackbone=Qwen3-Omni-30B-A3B-Instruct2026.05———36.3
ShiftDCBackbone=Qwen2.5-Omni-3B2026.05———46.3
ShiftDCBackbone=MiniCPM-o-4.52026.05———57.6
ShiftDCBackbone=Qwen3-Omni-30B-A3B-Instruct2026.05———26.9
SPA-VL-DPOBackbone=Qwen2.5-Omni-3B2026.05———67.4
VLGuard-SFTBackbone=Qwen2.5-Omni-3B2026.05———36.2