General Language Model Evaluation on WildBench
26.95WildBench ScorePUGC
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| PUGCModel=PUGC+DPO, Alignment=DPO2025.06 | 26.95 | 46.56 | 33.36 | 11.43 | 40.2 | 17.16 | |
| Mistral-7B-instructModel=Mistral-7B-instruct2025.06 | 25.63 | 42.07 | 30.06 | 10.08 | 40.1 | 18.4 |