List your company

Harmful Input Detection on JailbreakBench

100Accuracy (MCA)ReGA
45.9259.967488.04Jun 2, 2025

Evaluation results

MethodLinks
ReGAModel=qwen2025.0610092
ReGAModel=vicuna2025.069494
ReGAModel=mistral2025.069082
ReGAModel=baichuan2025.069090
ReGAModel=llama2025.068989
ReGAModel=Average2025.068582
ReGAModel=koala2025.064843