Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2602.04447 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866913100365561856 |
|---|---|
| author | Frisoni, Giacomo Molfetta, Lorenzo Freddi, Davide Moro, Gianluca |
| author_facet | Frisoni, Giacomo Molfetta, Lorenzo Freddi, Davide Moro, Gianluca |
| contents | Modern chess language models are dense transformers trained on millions of games played by thousands of high-rated individuals. However, these monolithic networks tend to collapse into mode-averaged behavior, where stylistic boundaries are blurred, and rare but effective strategies are suppressed. To counteract homogenization, we introduce Mixture-of-Masters (MoM), the first chess mixture-of-experts model with small-sized GPT experts emulating world-class grandmasters. For each move, a post-hoc learnable gating network selects the most appropriate persona to channel depending on the game state, allowing MoM to switch its style dynamically, e.g., Tal's offensive vocation or Petrosian's defensive solidity. When evaluated against Stockfish on unseen standard games, MoM outperforms both dense individual expert networks and popular GPT baselines trained on aggregated data, while ensuring generation variety, control, and interpretability. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_04447 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Mixture of Masters: Sparse Chess Language Models with Player Routing Frisoni, Giacomo Molfetta, Lorenzo Freddi, Davide Moro, Gianluca Machine Learning Artificial Intelligence Modern chess language models are dense transformers trained on millions of games played by thousands of high-rated individuals. However, these monolithic networks tend to collapse into mode-averaged behavior, where stylistic boundaries are blurred, and rare but effective strategies are suppressed. To counteract homogenization, we introduce Mixture-of-Masters (MoM), the first chess mixture-of-experts model with small-sized GPT experts emulating world-class grandmasters. For each move, a post-hoc learnable gating network selects the most appropriate persona to channel depending on the game state, allowing MoM to switch its style dynamically, e.g., Tal's offensive vocation or Petrosian's defensive solidity. When evaluated against Stockfish on unseen standard games, MoM outperforms both dense individual expert networks and popular GPT baselines trained on aggregated data, while ensuring generation variety, control, and interpretability. |
| title | Mixture of Masters: Sparse Chess Language Models with Player Routing |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2602.04447 |