A multi-agent LLM meta-learning system is introduced to co-evolve game-agnostic procedural search mechanisms in C++ alongside domain heuristics synthesized directly from game rules, achieving top-tier ratings and pairwise ballot majorities over most baselines across independent evolutionary runs.
Abstract
General Game Playing across arbitrary games from rules alone remains challenging due to differing algorithmic requirements across game classes and strict decision-time constraints. Rather than hand-designing search heuristics for specific domains, can we leverage Large Language Models (LLMs) to discover general game-playing algorithms? Because language models can propose and refactor structured code, they provide an expressive proposal engine for exploring the space of algorithmic designs. We introduce a multi-agent LLM meta-learning system to co-evolve game-agnostic procedural search mechanisms in C++ alongside domain heuristics synthesized directly from game rules. Controlling the compute budget, we benchmark the discovered mechanisms across more than 400 diverse environments, including OpenSpiel training and held-out games, procedural simulation engines, and games with deep neural policy-value representations trained via PPO. Evaluated via AlphaRank stationary distributions and Soft Condorcet Optimization (SCO) against 15 established MCTS baselines, the discovered search mechanisms consistently achieve top-tier ratings and pairwise ballot majorities over most baselines across independent evolutionary runs, generalizing to unseen human-designed and procedurally synthesized games and remaining competitive with baselines on frozen neural network representations.
Recent advances in Large Language Models (LLMs) have demonstrated remarkable capabilities in web front-end execution, with browser-based game generation emerging as a particularly prominent frontier. While previous efforts frequently rely on complex multi-turn workflows or focus on static game evaluation benchmarks, th...
Haoyue Yang, Jing-Yao Li, Zhen Wu et al.· 0 citations
The Abstraction Agent is proposed, a zero-shot pipeline that uses a large language model (LLM) to discover continuous strategic features from a natural-language game description, score private states on these features, and cluster them into abstraction buckets, without any game-specific evaluator, training data, or gam...
Adversarial games have driven advances from heuristic search to reinforcement learning, yet learning and adapting strategies from limited samples remain challenging. AI agents offer an alternative by turning game experience into revisions of executable policies. Building on heuristic learning (HL), we formalize Adversa...
Kai-Sen Yang, Qing-Le Liu, Ke-Jin Wang et al.· 0 citations
Recent large language models (LLMs) can operate as coding agents that build complete games from natural language requests. Game development is especially demanding because program logic, visual and audio content, interfaces, interaction and playability must function together in one executable artifact. Measuring this c...
Kun Chen, Hao-Rong Hong, Pei-Zhong Gao et al.· 6 citations
This work fine-tunes a 4-bit quantized LLaMA-3 8B model with Low-Rank Adaptation on 256 curated FIREBALL-derived examples and suggests that parameter-efficient adaptation improves grounding, while formatting, argument, and reference errors remain common.
Ez Zwall, Kedong Chen, Ya-Lun Zhou et al.· International Conference on...· 0 citations
Through PlayTrain, an RL framework that combines the abilities of large language models to robustly generate JavaScript games from a minimal human prompt, and an efficient pipeline that can run any JS game in a standard'gym'environment, this work reimagine RL VGE development.
R. Truong, Lance Ying, Samuel Gershman et al.· 0 citations