Review Summarization on Persona-based Review Summarization
66.8Overall ScoreRL
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| RLJudge=Qwen3-235B2025.12 | 66.8 | 68.4 | 42.9 | 89.2 | |
| RLJudge=gpt-oss-120b2025.12 | 63 | 62.9 | 64.2 | 62 | |
| SFTJudge=Qwen3-235B2025.12 | 50.7 | 44 | 53.1 | 55.1 | |
| SFTJudge=gpt-oss-120b2025.12 | 49 | 48.8 | 53.2 | 45.2 | |
| IPTJudge=Qwen3-235B2025.12 | 48.9 | 43.5 | 70.9 | 32.2 | |
| BaseJudge=gpt-oss-120b2025.12 | 44.1 | 46.4 | 36.1 | 49.7 | |
| IPTJudge=gpt-oss-120b2025.12 | 43.9 | 41.9 | 46.6 | 43.1 | |
| BaseJudge=Qwen3-235B2025.12 | 33.6 | 44.1 | 33.1 | 23.5 |