Resource-constrained Amazons chess decision framework integrating large language models and graph attention

TL;DR

This paper introduces a resource-efficient hybrid AI framework for the Amazons game, combining graph-based reasoning and large language models to improve decision accuracy with limited computational resources.

Contribution

It presents a novel lightweight hybrid approach integrating graph attention, genetic algorithms, and GPT-4o-mini to enhance game decision-making under resource constraints.

Findings

Achieved 15-56% improvement in decision accuracy over baselines.

Outperformed its teacher model GPT-4o-mini in win rate at low node counts.

Demonstrated high decision accuracy with significantly fewer computational resources.

Abstract

Artificial intelligence has advanced significantly through the development of intelligent game-playing systems, providing rigorous testbeds for decision-making, strategic planning, and adaptive learning. However, resource-constrained environments pose critical challenges, as conventional deep learning methods heavily rely on extensive datasets and computational resources. In this paper, we propose a lightweight hybrid framework for the Game of the Amazons, which explores the paradigm of weak-to-strong generalization by integrating the structural reasoning of graph-based learning with the generative capabilities of large language models. Specifically, we leverage a Graph Attention Autoencoder to inform a multi-step Monte Carlo Tree Search, utilize a Stochastic Graph Genetic Algorithm to optimize evaluation signals, and harness GPT-4o-mini to generate synthetic training data. Unlike traditional approaches that rely on expert demonstrations, our framework learns from noisy and imperfect supervision. We demonstrate that the Graph Attention mechanism effectively functions as a structural filter, denoising the LLM's outputs. Experiments on a 10$\times$10 Amazons board show that our hybrid approach not only achieves a 15\%--56\% improvement in decision accuracy over baselines but also significantly outperforms its teacher model (GPT-4o-mini), achieving a competitive win rate of 45.0\% at N=30 nodes and a decisive 66.5\% at only N=50 nodes. These results verify the feasibility of evolving specialized, high-performance game AI from general-purpose foundation models under stringent computational constraints.