List your company

Jailbreak Detection on GCG

99AccuracyIMAG
-3.9622.7749.576.23Dec 3, 2025

Evaluation results

MethodLinks
IMAGTarget LLM=Vicuna-13B2025.129999
IMAGTarget LLM=Llama2-7B2025.129999
IMAGTarget LLM=Mistral-7B2025.129899
IMAGTarget LLM=Llama3-8B2025.129899
GradSafeTarget LLM=Llama2-7B2025.129798
IMAGTarget LLM=Vicuna-7B2025.129698
PPLTarget LLM=Vicuna-13B2025.127986
PPLTarget LLM=Llama2-7B2025.127986
LlamaGTarget LLM=Mistral-7B2025.127887
PPLTarget LLM=Llama3-8B2025.127786
LlamaGTarget LLM=Vicuna-13B2025.127686
LlamaGTarget LLM=Vicuna-7B2025.127586
GradSafeTarget LLM=Mistral-7B2025.126377
LlamaGTarget LLM=Llama3-8B2025.125470
Self-ExTarget LLM=Mistral-7B2025.125269
PPLTarget LLM=Vicuna-7B2025.124760
PPLTarget LLM=Mistral-7B2025.123348
LlamaGTarget LLM=Llama2-7B2025.123248
Self-ExTarget LLM=Llama2-7B2025.122832
OAPITarget LLM=Mistral-7B2025.121323
OAPITarget LLM=Llama3-8B2025.121221
Self-ExTarget LLM=Llama3-8B2025.121221
OAPITarget LLM=Vicuna-7B2025.121018
OAPITarget LLM=Vicuna-13B2025.12816
OAPITarget LLM=Llama2-7B2025.12714
Self-ExTarget LLM=Vicuna-7B2025.1200
GradSafeTarget LLM=Vicuna-7B2025.1200
Self-ExTarget LLM=Vicuna-13B2025.1200
GradSafeTarget LLM=Vicuna-13B2025.1200
GradSafeTarget LLM=Llama3-8B2025.1200