让棋类模型像大师一样切换风格，动态选择最佳下法策略。
Mixture of Masters: Sparse Chess Language Models with Player Routing
- 用小型GPT专家模拟大师风格，通过门控网络按局面选风格。
- 在未见对局中击败密集模型和主流GPT基线，胜率更高且风格多样。
- 适合研究棋艺风格建模、可解释生成与动态策略切换的场景。
现代国际象棋语言模型通常是在数百万盘由数千名高分玩家对弈的数据上训练的密集Transformer模型。然而，这类单体网络往往趋向于模式平均行为，导致风格边界模糊，以及罕见但高效的策略被压制。为应对同质化问题，我们提出首个棋类混合专家模型Mixture-of-Masters（MoM），采用小型GPT专家模拟世界级特级大师。每个走法由一个可学习的后验门控网络根据当前棋局状态选择最合适的专家人格，实现风格动态切换，例如塔尔的进攻性或佩特罗相的防守稳固性。在与Stockfish对抗未见标准对局的评估中，MoM优于单一密集专家网络及基于聚合数据训练的流行GPT基线，同时保证生成多样性、可控性与可解释性。
原文摘要 · Abstract (English)
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.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。