Safety Evaluation on HarmfulQ
1.5ASRQwQ
Evaluation results
| Method | Links | ||
|---|---|---|---|
| QwQConflict=direct_q2026.04 | 1.5 | — | |
| Llama-NemotronConflict=direct_q2026.04 | 2.5 | — | |
| Llama-NemotronConflict=inner2026.04 | 6.5 | 4 | |
| Llama-NemotronConflict=dilemma2026.04 | 20 | 17.5 | |
| QwQConflict=inner2026.04 | 33 | 31.5 | |
| QwQConflict=dilemma2026.04 | 39.5 | 38 |