Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

MM-SafetyBench

Benchmarks

Task NameDataset NameSOTA ResultTrend
Video JailbreakingMM-SafetyBench 1.0 (test)
Attack Success Rate96
48
Safety EvaluationMM-SafetyBench
Average ASR0
42
MLLM JailbreakingMM-SafetyBench Physical Harm scenario
ASR6
15
Multimodal Jailbreak DefenseMM-SafetyBench (full)
ASR (Illegal Activity - S)1.03
12
Multimodal Safety DefenseMM-SafetyBench SD_TYPO
Average ASR12
10
Multimodal Safety DefenseMM-SafetyBench SD
Average ASR0.09
10
Harmful Rate EvaluationMM-SafetyBench OCR (test)
Illegal Activity Rate0
10
Jailbreak DetectionMM-SafetyBench
AUROC99.18
9
Multimodal Safety EvaluationMM-SafetyBench SD + TYPO + SD_TYPO (test)
ASR Score0.08
8
Jailbreaking AttackMM-SafetyBench
Attack Success Rate (ASR)91.5
8
Multi-turn Jailbreaking AttackMM-SafetyBench Turn 2
ASR24.42
5
Safety EvaluationMM-SafetyBench SD 1.0
Illegal Activity Score48.3
5
Safety EvaluationMM-SafetyBench H
Safety Score1
4
Showing 13 of 13 rows