List your company

Multi-task Language Understanding on MMLU (pass@1 Accuracy)

88pass@1 AccuracySelf-Reminder
32.8847.1961.575.81Aug 6, 2025

Evaluation results

MethodLinks
Self-ReminderBackbone=R1-Qwen-32B2025.0888
No DefenseBackbone=R1-Qwen-32B2025.0887
RealSafe-R1Backbone=R1-Qwen-32B2025.0886
No DefenseBackbone=R1-Qwen-14B2025.0885
ThinkingIBackbone=R1-Qwen-14B2025.0885
SAFEPATH-ZSBackbone=R1-Qwen-32B2025.0885
Self-ReminderBackbone=R1-Qwen-14B2025.0884
SAFEPATH-ZSBackbone=R1-Qwen-14B2025.0884
SafeKeyBackbone=R1-Qwen-14B2025.0883
ReasoningGuardBackbone=R1-Qwen-32B2025.0883
ThinkingIBackbone=R1-Qwen-32B2025.0882
RealSafe-R1Backbone=R1-Qwen-14B2025.0881
ReasoningGuardBackbone=R1-Qwen-14B2025.0881
SmoothLLMBackbone=R1-Qwen-14B2025.0872
SmoothLLMBackbone=R1-Qwen-32B2025.0868
ParaphraseBackbone=R1-Qwen-32B2025.0838
ParaphraseBackbone=R1-Qwen-14B2025.0835