Conversational on MixEval
76ScoreQwen3.5-9B
Evaluation Results
| Method | Links | |
|---|---|---|
| Qwen3.5-9BParameters=9B, Variant=Thinking2026.05 | 76 | |
| Qwen3.5-4BParameters=4B, Variant=Thinking2026.05 | 71.9 | |
| Ministral-3-14BParameters=14B, Variant=Thinking2026.05 | 70.8 | |
| OLMo-3-7BParameters=7B, Variant=Thinking2026.05 | 67 | |
| Mellum 2 (RL)Post-training Stage=RL, Parameters=2.5B/12B, Variant=Thinking2026.05 | 66.9 | |
| Mellum 2 (SFT)Post-training Stage=SFT, Parameters=2.5B/12B, Variant=Thinking2026.05 | 63.4 |