Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Multimodal Safety Evaluation on FigStep
Loading...
0.21
ASR (%)
CrossGuard
-3.1048
19.2701
41.645
64.0199
Oct 20, 2025
ASR (%)
Updated 1mo ago
Evaluation Results
Method
Method
Links
ASR (%)
CrossGuard
Category=MLLM Guardrails
2025.10
0.21
GPT-4o
Category=Online MLLMs
2025.10
1.6
JailDAM
Category=MLLM Guardrails
2025.10
6
Claude-3.5-Sonnet
Category=Online MLLMs
2025.10
13
Qwen2.5-VL-7B
Category=Offline MLLMs
2025.10
24.2
LLaVA-1.5-7B (base)
Category=Offline MLLMs
2025.10
62.6
Llama-Guard3-Vision
Category=MLLM Guardrails
2025.10
66.92
HiddenDetect
Category=MLLM Guardrails
2025.10
72.2
LlavaGuard
Category=MLLM Guardrails
2025.10
83.08
Feedback
Search any
task
Search any
task