Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
VLM Jailbreak Detection on VLM Jailbreak Average (MM-SafetyBench, JailBreakV-28K, FigImg)
Loading...
0.94
Average AUROC
MTK
0.53128
0.63739
0.7435
0.84961
Jun 5, 2026
Average AUROC
Updated 1mo ago
Evaluation Results
Method
Method
Links
Average AUROC
MTK
VLM=LLaVa
2026.06
0.94
HiddenDetect
VLM=LLaVa
2026.06
0.928
MTK
VLM=Qwen-VL
2026.06
0.924
HiddenDetect
VLM=Qwen-VL
2026.06
0.878
ECSO
VLM=LLaVa
2026.06
0.825
JailGuard
VLM=Qwen-VL
2026.06
0.809
MirrorCheck
VLM=Qwen-VL
2026.06
0.708
ECSO
VLM=Qwen-VL
2026.06
0.683
MirrorCheck
VLM=LLaVa
2026.06
0.669
CIDER
VLM=LLaVa
2026.06
0.655
CIDER
VLM=Qwen-VL
2026.06
0.655
JailDAM
VLM=LLaVa
2026.06
0.627
JailDAM
VLM=Qwen-VL
2026.06
0.593
JailGuard
VLM=LLaVa
2026.06
0.547
Feedback
Search any
task
Search any
task