List your company

Malicious Prompt Detection on VLGuard & MLLMGuard π = 0.005

94.37AUROCVLMGuard
49.587661.213872.8484.4662Oct 1, 2024

Evaluation results

MethodLinks
VLMGuardBackbone=Qwen, Type=Activation, Single inference=true, Single LM=true2024.1094.3794.79
VLMGuardBackbone=LLaVA, Type=Activation, Single inference=true, Single LM=true2024.1091.9493
HiddenDetectBackbone=LLaVA, Type=Activation, Single inference=true, Single LM=true2024.1088.8788.53
GPT-5.4Backbone=N/A, Type=LLM, Single inference=true, Single LM=false2024.1086.4189.4
ASTRABackbone=LLaVA, Type=Activation, Single inference=true, Single LM=true2024.1082.1183.86
Qwen2.5-VL-72B-InstructBackbone=N/A, Type=LLM, Single inference=true, Single LM=false2024.1081.2785.33
ASTRABackbone=Qwen, Type=Activation, Single inference=true, Single LM=true2024.1080.5282.7
InternVL3-78B-InstructBackbone=N/A, Type=LLM, Single inference=true, Single LM=false2024.1078.2982.51
GradSafeBackbone=Qwen, Type=Uncertainty, Single inference=true, Single LM=true2024.1076.8778.32
GradSafeBackbone=LLaVA, Type=Uncertainty, Single inference=true, Single LM=true2024.1076.1277.1
Ovis2-34BBackbone=N/A, Type=LLM, Single inference=true, Single LM=false2024.1073.7871.4
Gradient CuffBackbone=Qwen, Type=Uncertainty, Single inference=false, Single LM=true2024.1073.5780.56
Gradient CuffBackbone=LLaVA, Type=Uncertainty, Single inference=false, Single LM=true2024.1071.7778.55
HiddenDetectBackbone=Qwen, Type=Activation, Single inference=true, Single LM=true2024.1070.3476.57
Self-detectionBackbone=Qwen, Type=LLM, Single inference=true, Single LM=true2024.1065.0772.52
Self-detectionBackbone=LLaVA, Type=LLM, Single inference=true, Single LM=true2024.1063.7771.14
LLaVAGuardBackbone=N/A, Type=LLM, Single inference=true, Single LM=false2024.1062.7559.62
LlamaGuard3-VisionBackbone=N/A, Type=LLM, Single inference=true, Single LM=false2024.1061.4169.57
JailGuardBackbone=LLaVA, Type=Mutation, Single inference=false, Single LM=true2024.1061.0861.34
CIDERBackbone=LLaVA, Type=Denoising, Single inference=true, Single LM=false2024.1061.0559.99
JailGuardBackbone=Qwen, Type=Mutation, Single inference=false, Single LM=true2024.1060.5162.15
CIDERBackbone=Qwen, Type=Denoising, Single inference=true, Single LM=false2024.1059.2759.44
PerplexityBackbone=Qwen, Type=Uncertainty, Single inference=true, Single LM=true2024.1054.1963.68
MirrorCheckBackbone=Qwen, Type=Denoising, Single inference=true, Single LM=false2024.1052.6857.25
PerplexityBackbone=LLaVA, Type=Uncertainty, Single inference=true, Single LM=true2024.1052.3161.4
MirrorCheckBackbone=LLaVA, Type=Denoising, Single inference=true, Single LM=false2024.1051.3154.8