List your company

Harmfulness Evaluation on Mousetrap

3.78Harmfulness ScoreSelf-Reminder
0.90961.65482.43.1452Aug 6, 2025

Evaluation results

MethodLinks
Self-ReminderModel=R1-Llama-8B2025.083.78
ParaphraseModel=R1-Llama-8B2025.083.52
No DefenseModel=R1-Llama-8B2025.083.2
SafeKeyModel=R1-Llama-8B2025.083.14
SafeDecodingModel=R1-Llama-8B2025.083.12
SAFEPATH-ZSModel=R1-Llama-8B2025.082.9
ParaphraseModel=R1-Qwen-7B2025.082.82
SafeKeyModel=R1-Qwen-7B2025.082.72
SmoothLLMModel=R1-Llama-8B2025.082.72
No DefenseModel=R1-Qwen-7B2025.082.7
SafeDecodingModel=R1-Qwen-7B2025.082.66
SmoothLLMModel=R1-Qwen-7B2025.082.48
Self-ReminderModel=R1-Qwen-7B2025.082.32
RealSafe-R1Model=R1-Qwen-7B2025.081.92
SAFEPATH-ZSModel=R1-Qwen-7B2025.081.8
ThinkingIModel=R1-Qwen-7B2025.081.78
ThinkingIModel=R1-Llama-8B2025.081.66
RealSafe-R1Model=R1-Llama-8B2025.081.46
ReasoningGuardModel=R1-Llama-8B2025.081.34
SAFEPATH-FTModel=R1-Qwen-7B2025.081.3
ReasoningGuardModel=R1-Qwen-7B2025.081.14
SAFEPATH-FTModel=R1-Llama-8B2025.081.02