List your company

Safety Evaluation on XSTest Safe

4FCReasoningGuard
0.1626.085277.92Aug 6, 2025

Evaluation results

MethodLinks
ReasoningGuardModel=R1-Qwen-7B, Output Component=Reasoning2025.084294
ThinkingIModel=R1-Llama-8B, Output Component=Reasoning2025.0868212
ThinkingIModel=R1-Qwen-7B, Output Component=Reasoning2025.0885834
SAFEPATH-FTModel=R1-Llama-8B, Output Component=Reasoning2025.08285022
RealSafe-R1Model=R1-Qwen-7B, Output Component=Reasoning2025.08321058
SAFEPATH-FTModel=R1-Qwen-7B, Output Component=Reasoning2025.08322840
ReasoningGuardModel=R1-Qwen-14B, Evaluation Component=Reasoning2025.0832068
RealSafe-R1Model=R1-Qwen-7B, Output Component=Final Answer2025.08344026
ReasoningGuardModel=R1-Qwen-32B, Evaluation Component=Reasoning2025.0834264
ReasoningGuardModel=R1-Llama-8B, Output Component=Reasoning2025.0838062
SAFEPATH-ZSModel=R1-Qwen-7B, Output Component=Reasoning2025.0840852
SAFEPATH-FTModel=R1-Llama-8B, Output Component=Final Answer2025.08404020
ThinkingIModel=R1-Qwen-7B, Output Component=Final Answer2025.0842508
SAFEPATH-ZSModel=R1-Llama-8B, Output Component=Reasoning2025.08441046
RealSafe-R1Model=R1-Llama-8B, Output Component=Reasoning2025.08461044
RealSafe-R1Model=R1-Llama-8B, Output Component=Final Answer2025.08502426
SmoothLLMModel=R1-Qwen-7B, Output Component=Reasoning2025.0852840
ReasoningGuardModel=R1-Qwen-7B, Output Component=Final Answer2025.08543610
SafeKeyModel=R1-Qwen-7B, Output Component=Reasoning2025.0860832
SafeKeyModel=R1-Llama-8B, Output Component=Reasoning2025.0860634
ThinkingIModel=R1-Llama-8B, Output Component=Final Answer2025.08603010
ThinkingIModel=R1-Qwen-32B, Evaluation Component=Reasoning2025.08603010
SAFEPATH-FTModel=R1-Qwen-7B, Output Component=Final Answer2025.0862308
Self-ReminderModel=R1-Qwen-7B, Output Component=Reasoning2025.0862038
SAFEPATH-ZSModel=R1-Qwen-14B, Evaluation Component=Reasoning2025.0864036
SmoothLLMModel=R1-Llama-8B, Output Component=Reasoning2025.0866430
RealSafe-R1Model=R1-Qwen-14B, Evaluation Component=Reasoning2025.0868230
SAFEPATH-ZSModel=R1-Qwen-7B, Output Component=Final Answer2025.08701416
SafeKeyModel=R1-Qwen-7B, Output Component=Final Answer2025.08721414
ThinkingIModel=R1-Qwen-14B, Evaluation Component=Reasoning2025.08721810
SAFEPATH-ZSModel=R1-Qwen-32B, Evaluation Component=Reasoning2025.0872424
ParaphraseModel=R1-Qwen-7B, Output Component=Reasoning2025.0874422
SmoothLLMModel=R1-Qwen-7B, Output Component=Final Answer2025.0874818
RealSafe-R1Model=R1-Qwen-14B, Evaluation Component=Final Answer2025.0874224
RealSafe-R1Model=R1-Qwen-32B, Evaluation Component=Reasoning2025.0874224
SmoothLLMModel=R1-Llama-8B, Output Component=Final Answer2025.0876816
SafeKeyModel=R1-Qwen-14B, Evaluation Component=Reasoning2025.0876420
ReasoningGuardModel=R1-Qwen-32B, Evaluation Component=Final Answer2025.0878616
SAFEPATH-ZSModel=R1-Llama-8B, Output Component=Final Answer2025.0880812
ReasoningGuardModel=R1-Llama-8B, Output Component=Final Answer2025.0880812
Self-ReminderModel=R1-Qwen-7B, Output Component=Final Answer2025.0882612
No DefenseModel=R1-Llama-8B, Output Component=Reasoning2025.0882018
Self-ReminderModel=R1-Llama-8B, Output Component=Reasoning2025.0882018
SmoothLLMModel=R1-Qwen-14B, Evaluation Component=Reasoning2025.0882018
SafeKeyModel=R1-Llama-8B, Output Component=Final Answer2025.0884412
SmoothLLMModel=R1-Qwen-32B, Evaluation Component=Reasoning2025.0884016
Self-ReminderModel=R1-Qwen-14B, Evaluation Component=Reasoning2025.0886014
RealSafe-R1Model=R1-Qwen-32B, Evaluation Component=Final Answer2025.088686
Self-ReminderModel=R1-Qwen-32B, Evaluation Component=Reasoning2025.0886014
SmoothLLMModel=R1-Qwen-14B, Evaluation Component=Final Answer2025.0888012
ThinkingIModel=R1-Qwen-14B, Evaluation Component=Final Answer2025.088848
SmoothLLMModel=R1-Qwen-32B, Evaluation Component=Final Answer2025.0888210
No DefenseModel=R1-Qwen-7B, Output Component=Reasoning2025.089028
SafeDecodingModel=R1-Qwen-7B, Output Component=Reasoning2025.0890010
ReasoningGuardModel=R1-Qwen-14B, Evaluation Component=Final Answer2025.089082
No DefenseModel=R1-Qwen-32B, Evaluation Component=Reasoning2025.0890010
ParaphraseModel=R1-Qwen-7B, Output Component=Final Answer2025.089244
SafeDecodingModel=R1-Llama-8B, Output Component=Reasoning2025.089208
SAFEPATH-ZSModel=R1-Qwen-14B, Evaluation Component=Final Answer2025.089208
ParaphraseModel=R1-Qwen-32B, Evaluation Component=Reasoning2025.089208
ThinkingIModel=R1-Qwen-32B, Evaluation Component=Final Answer2025.089226
No DefenseModel=R1-Qwen-14B, Evaluation Component=Reasoning2025.089406
SafeKeyModel=R1-Qwen-14B, Evaluation Component=Final Answer2025.089442
SAFEPATH-ZSModel=R1-Qwen-32B, Evaluation Component=Final Answer2025.089424
No DefenseModel=R1-Qwen-7B, Output Component=Final Answer2025.089622
ParaphraseModel=R1-Llama-8B, Output Component=Reasoning2025.089604
Self-ReminderModel=R1-Qwen-14B, Evaluation Component=Final Answer2025.089604
No DefenseModel=R1-Qwen-32B, Evaluation Component=Final Answer2025.089604
Self-ReminderModel=R1-Qwen-32B, Evaluation Component=Final Answer2025.089604
SafeDecodingModel=R1-Qwen-7B, Output Component=Final Answer2025.089802
No DefenseModel=R1-Llama-8B, Output Component=Final Answer2025.089802
SafeDecodingModel=R1-Llama-8B, Output Component=Final Answer2025.089802
ParaphraseModel=R1-Llama-8B, Output Component=Final Answer2025.089802
Self-ReminderModel=R1-Llama-8B, Output Component=Final Answer2025.089802
ParaphraseModel=R1-Qwen-14B, Evaluation Component=Reasoning2025.089802
ParaphraseModel=R1-Qwen-14B, Evaluation Component=Final Answer2025.089802
ParaphraseModel=R1-Qwen-32B, Evaluation Component=Final Answer2025.089802
No DefenseModel=R1-Qwen-14B, Evaluation Component=Final Answer2025.0810000