ChessMimic: 用于在线闪电棋中人类走棋、时钟和结果预测的按等级划分的Transformer模型
ChessMimic: Per-Rating Transformer Models for Human Move, Clock, and Outcome Prediction in Online Blitz Chess
AI总结：
提出ChessMimic系统，包含三个小型编码器Transformer模型，分别用于走棋、思考时间和结果预测，通过按Elo等级分段训练实现更精细的技能校准，在Lichess闪电棋数据上走棋预测准确率超越Maia-2，结果预测AUC达0.78，时钟模型提供可用但非最优的思考时间信号。
摘要：
我们提出了ChessMimic，一个由三个小型编码器Transformer组成的系统——分别用于走棋、思考时间和结果预测——以局面、最近走棋历史、玩家等级和时钟状态为条件。我们为每100 Elo等级区间拟合每个模型的独立实例，以参数效率换取更精细的技能校准。在Lichess Rated Blitz游戏的一个月保留切片上，ChessMimic的人类走棋预测准确率在每个Elo区间都优于Maia-2。与Maia-3相比，我们的9M参数模型的准确率介于Maia-3-5M和Maia-3-23M之间，且没有几何注意力偏置的额外复杂性。除了走棋匹配模型，我们还训练了一个游戏结果模型，该模型不仅以局面为条件，还以玩家等级、时间控制和剩余时钟时间为条件。结果模型在样本外达到了0.78的AUC，击败了Maia-2以及基于子力、等级和时钟时间的逻辑回归。最后，我们训练了一个时钟模型来预测人类思考时间。该时钟模型在ALLIE风格过滤器下提供了可用但非最优的每步思考时间信号（Pearson r = 0.41，Spearman rho = 0.50，MAE 4.10秒，而ALLIE报告的r = 0.70），残差差距集中在每位置桶的锐度上，而非桶边际校准。公开演示在1e4.ai，我们在GitHub上发布了代码、每个区间的权重以及C++数据过滤管道代码。
英文摘要：
We present ChessMimic, a system of three small encoder-only transformers - for move, thinking-time, and outcome prediction - conditioned on the position, recent move history, player rating, and clock state. We fit a separate instance of each model per 100-Elo rating band, trading parameter efficiency for sharper per-skill calibration. On a held-out month-wide slice of Lichess Rated Blitz games ChessMimic's human move prediction accuracy outperforms Maia-2 in every Elo band. Compared to Maia-3, our 9M parameter model's accuracy sits between Maia-3-5M and Maia-3-23M without the additional complexity of Geometric Attention Bias. In addition to the move matching model, we also train a game outcome model that conditions not only on the position, but also player ratings, time control, and remaining clock times. The outcome model achieves an AUC of 0.78 out of sample, beating Maia-2 as well as logistic regressions based on material, ratings, and clock time. Finally, we train a clock model that predicts human thinking times. The clock model provides a usable but non-SOTA per-ply think-time signal under ALLIE-style filters (Pearson r = 0.41, Spearman rho = 0.50, MAE 4.10 s, against ALLIE's reported r = 0.70), with the residual gap concentrated in per-position bucket sharpness rather than bucket-marginal calibration. A public demo is at 1e4.ai and we release code, per-band weights, and the C++ data-filter pipeline code in GitHub.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
