Mixing Expert Knowledge: Bring Human Thoughts Back To the Game of Go

TL;DR

This paper introduces LoGos, a large language model fine-tuned with Go expertise and reasoning data, achieving human professional-level Go gameplay and demonstrating the integration of expert domain knowledge with general reasoning in LLMs.

Contribution

The paper presents a novel fine-tuning and reinforcement learning approach to incorporate expert Go knowledge into LLMs, enabling high-level Go gameplay performance.

Findings

LoGos matches human professional Go players in performance.

LoGos surpasses all existing LLMs in Go gameplay.

First LLM to reach human professional-level in Go.

Abstract

Large language models (LLMs) have demonstrated exceptional performance in reasoning tasks such as mathematics and coding, matching or surpassing human capabilities. However, these impressive reasoning abilities face significant challenges in specialized domains. Taking Go as an example, although AlphaGo has established the high performance ceiling of AI systems in Go, mainstream LLMs still struggle to reach even beginner-level proficiency, let alone perform natural language reasoning. This performance gap between general-purpose LLMs and domain experts is significantly limiting the application of LLMs on a wider range of domain-specific tasks. In this work, we aim to bridge the divide between LLMs' general reasoning capabilities and expert knowledge in domain-specific tasks. We perform mixed fine-tuning with structured Go expertise and general long Chain-of-Thought (CoT) reasoning data as a cold start, followed by reinforcement learning to integrate expert knowledge in Go with general reasoning capabilities. Through this methodology, we present \textbf{LoGos}, a powerful LLM that not only maintains outstanding general reasoning abilities, but also conducts Go gameplay in natural language, demonstrating effective strategic reasoning and accurate next-move prediction. LoGos achieves performance comparable to human professional players, substantially surpassing all existing LLMs. Through this work, we aim to contribute insights on applying general LLM reasoning capabilities to specialized domains. We will release the first large-scale Go dataset for LLM training, the first LLM Go evaluation benchmark, and the first general LLM that reaches human professional-level performance in Go at: https://github.com/Entarochuan/LoGos.