Board-game self-play on Hex 10k
96Best-Response Win RateMMD
Evaluation results
| Method | Links | |
|---|---|---|
| MMDevaluation_protocol=vs. the frozen policy, aggregation=mean±std over 8 seeds2026.06 | 96 | |
| fict.evaluation_protocol=vs. the frozen policy, aggregation=mean±std over 8 seeds2026.06 | 96 | |
| naiveevaluation_protocol=vs. the frozen policy, aggregation=mean±std over 8 seeds2026.06 | 93 | |
| GARIPevaluation_protocol=vs. the frozen policy, aggregation=mean±std over 8 seeds2026.06 | 92 | |
| R-NaDevaluation_protocol=vs. the frozen policy, aggregation=mean±std over 8 seeds2026.06 | 89 |