Over-refusal Assessment on XS (test)
93.1F1 ScoreSelf-Reminder
Evaluation Results
| Method | Links | |
|---|---|---|
| Self-ReminderModel Scale=Qwen3-14B2026.01 | 93.1 | |
| Self-ReminderModel Scale=Qwen3-8B2026.01 | 92.6 | |
| Self-ReminderModel Scale=Qwen3-4B2026.01 | 92.5 | |
| ReasoningGuardModel Scale=Qwen3-4B2026.01 | 92.4 | |
| Self-GuardModel Scale=Qwen3-8B2026.01 | 92 | |
| ReasoningGuardModel Scale=Qwen3-8B2026.01 | 91.3 | |
| SafeKeyModel Scale=Qwen3-14B2026.01 | 91 | |
| Self-GuardModel Scale=Qwen3-14B2026.01 | 91 | |
| STAR-1Model Scale=Qwen3-14B2026.01 | 89.6 | |
| Self-GuardModel Scale=Qwen3-4B2026.01 | 89.3 | |
| SafeKeyModel Scale=Qwen3-8B2026.01 | 89.3 | |
| STAR-1Model Scale=Qwen3-8B2026.01 | 88.4 | |
| Alpha-steerModel Scale=Qwen3-14B2026.01 | 87.4 | |
| InstructModel Scale=Qwen3-14B2026.01 | 87.1 | |
| ReasoningGuardModel Scale=Qwen3-14B2026.01 | 86.7 | |
| Alpha-steerModel Scale=Qwen3-4B2026.01 | 86.3 | |
| InstructModel Scale=Qwen3-4B2026.01 | 86.1 | |
| STAR-1Model Scale=Qwen3-4B2026.01 | 85.2 | |
| Alpha-steerModel Scale=Qwen3-8B2026.01 | 84.5 | |
| SafeKeyModel Scale=Qwen3-4B2026.01 | 84.1 | |
| InstructModel Scale=Qwen3-8B2026.01 | 83.1 | |
| SafeChainModel Scale=Qwen3-14B2026.01 | 69.1 | |
| SafeChainModel Scale=Qwen3-8B2026.01 | 63.5 | |
| SafeChainModel Scale=Qwen3-4B2026.01 | 51.8 |