List your company

Jailbreak Robustness on AutoRAN

0Harmfulness RateRealSafe-R1
-0.241.3834.62Aug 6, 2025

Evaluation results

MethodLinks
RealSafe-R1Model=R1-Qwen-14B2025.080
SafeKeyModel=R1-Qwen-14B2025.080
ThinkingIModel=R1-Qwen-14B2025.080
SAFEPATH-ZSModel=R1-Qwen-14B2025.080
ReasoningGuardModel=R1-Qwen-14B2025.080
No DefenseModel=R1-Qwen-32B2025.080
RealSafe-R1Model=R1-Qwen-32B2025.080
SAFEPATH-ZSModel=R1-Qwen-32B2025.080
ReasoningGuardModel=R1-Qwen-32B2025.080
No DefenseModel=R1-Qwen-14B2025.082
Self-ReminderModel=R1-Qwen-14B2025.082
SmoothLLMModel=R1-Qwen-14B2025.082
ParaphraseModel=R1-Qwen-32B2025.082
Self-ReminderModel=R1-Qwen-32B2025.084
SmoothLLMModel=R1-Qwen-32B2025.084
ParaphraseModel=R1-Qwen-14B2025.086
ThinkingIModel=R1-Qwen-32B2025.086