Review Generation on Board Game Playtesting Dataset
99.46FactualityGPT-5.1
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| GPT-5.12026.01 | 99.46 | 0.6934 | 4.26 | |
| Qwen3-235B2026.01 | 98.95 | 0.6572 | 3.56 | |
| MeepleLM2026.01 | 98.86 | 0.7117 | 4.34 | |
| Gemini3-Pro2026.01 | 98.28 | 0.648 | 3.98 | |
| Qwen3-8B2026.01 | 97.88 | 0.5936 | 1.58 | |
| MeepleLMAblation=w/o Persona2026.01 | 92.13 | 0.6771 | 3.56 | |
| MeepleLMAblation=w/o MDA2026.01 | 91.56 | 0.685 | 3.7 | |
| MeepleLMAblation=w/o Rulebook2026.01 | 59.87 | 0.697 | 3.3 |