paper-with-me

Papers

UniMaia: Steering Chess Policies with Language for Human-like Play

2026-05-26 · Sherman Siu, Lesley Istead arxiv

Recent advances in large language models have enabled natural language to serve as a flexible interface for controlling complex systems, but often at the cost of large-scale multimodal training or weakened domain-specific inductive biases. In structured decision-making domains such as chess, specialized policy networks achieve strong performance but lack semantic controllability, while prompt-conditioned language models are more flexible yet typically exhibit weaker domain grounding. We propose $\textbf{UniMaia}$, a framework for prompt-conditioned policy modulation that adapts a frozen Lc0-based chess policy network using a parameter-efficient text encoder and a ControlNet-style conditioning mechanism. UniMaia enables semantic control over gameplay, including opening selection and player strength, while preserving the pretrained policy representations. We further introduce $\textbf{UniMaia-Aux}$, which incorporates auxiliary temporal conditioning and behavioral prediction objectives. To support this work, we construct a large-scale metadata-augmented Lichess dataset, develop a semi-automated prompt-generation pipeline, and introduce benchmarks spanning both prompt-conditioned and metadata-conditioned settings. UniMaia achieves state-of-the-art expected accuracy on several prompt-conditioned benchmarks and competitive top-move accuracy on general instruction-following tasks, while remaining competitive with dedicated metadata-conditioned approaches on human move prediction benchmarks. UniMaia-Aux further improves expected accuracy and behavioral modeling across several evaluation settings, with modest trade-offs in top-move accuracy. Overall, our results demonstrate that prompt-conditioned control of domain-specific policy networks is feasible without end-to-end multimodal training, while highlighting trade-offs between controllability and predictive performance.

📄 PDF Abstract BibTeX arXiv:2605.27767

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Policy Gradient Steering: Interventions from Behavioral Objectives

2026-07-30 · Yoann Poupart, Aurélie Beynier, Nicolas Maudet arxiv

Activation steering has emerged in large language models as a lightweight alternative for dynamically changing a model's behavior at inference time. However, we show that existing steering methods fail to steer even a si…

Reinforcement Learning

Bridging the Gap between Expert and Language Models: Concept-guided Chess Commentary Generation and Evaluation

2024-10-28 · Jaechang Kim, Jinmin Goh, Inseok Hwang, Jaewoong Cho 외

Deep learning-based expert models have reached superhuman performance in decision-making domains such as chess and Go. However, it is under-explored to explain or comment on given decisions although it is important for h…

Decision MakingInformativeness

Improving Chess Commentaries by Combining Language Models with Symbolic Reasoning Engines

2022-12-15 · Andrew Lee, David Wu, Emily Dinan, Mike Lewis

Despite many recent advancements in language modeling, state-of-the-art language models lack grounding in the real world and struggle with tasks involving complex reasoning. Meanwhile, advances in the symbolic reasoning …

Language ModelingLanguage Modelling

ChessGPT: Bridging Policy Learning and Language Modeling

2023-06-15 · NeurIPS 2023 11 · Xidong Feng, Yicheng Luo, Ziyan Wang, Hongrui Tang 외

When solving decision-making tasks, humans typically depend on information from two key sources: (1) Historical policy data, which provides interaction replay from the environment, and (2) Analytical insights in natural …

Decision MakingLanguage ModelingLanguage Modelling

Explore the Reasoning Capability of LLMs in the Chess Testbed

2024-11-11 · Shu Wang, Lei Ji, Renxi Wang, Wenxiao Zhao 외

Reasoning is a central capability of human intelligence. In recent years, with the advent of large-scale datasets, pretrained large language models have emerged with new capabilities, including reasoning. However, these …