List your company

Jailbreaking Attack Detection on Technical Terms Jailbreaking Prompts

59Accuracy (MCA)ReGA
11.1623.583648.42Jun 2, 2025

Evaluation results

MethodLinks
ReGATarget LLM=vicuna2025.065939
ReGATarget LLM=mistral2025.065241
ReGATarget LLM=llama2025.065050
ReGATarget LLM=qwen2025.063922
ReGATarget LLM=baichuan2025.062626
ReGATarget LLM=koala2025.061313