用进化优化提升AI玩2048的策略，单智能体系统表现显著优于双智能体。
Merge and Conquer: Evolutionarily Optimizing AI for 2048
- 单智能体通过优化价值函数结合蒙特卡洛搜索，实现策略迭代。
- 平均每轮得分提升473.2分，训练周期间相关性达ρ=0.607。
- 适合研究动态环境下的决策优化与强化学习机制。
优化人工智能在动态环境中的表现仍是机器学习研究的核心挑战。本文以2048这一融合策略与随机性的二维滑动拼图游戏为实验场景，研究其决策制定、长期规划与动态适应能力。我们构建了两种系统：一种是双智能体元提示系统，由“思考者”大型语言模型（LLM）优化“执行者”LLM的策略；另一种是基于有限蒙特卡洛树搜索的价值函数优化的单智能体系统，并引入回滚机制防止性能退化。结果表明，进化优化在非确定性环境中具有潜力。单智能体系统取得显著改进，平均每轮得分提升473.2分，且训练周期间呈现明显上升趋势（相关系数ρ=0.607）。同时，LLM对游戏的理解持续深化，展现出更高级策略。而双智能体系统未见显著提升，揭示了元提示方法的内在局限。
原文摘要 · Abstract (English)
Optimizing artificial intelligence (AI) for dynamic environments remains a fundamental challenge in machine learning research. In this paper, we examine evolutionary training methods for optimizing AI to solve the game 2048, a 2D sliding puzzle. 2048, with its mix of strategic gameplay and stochastic elements, presents an ideal playground for studying decision-making, long-term planning, and dynamic adaptation. We implemented two distinct systems: a two-agent metaprompting system where a "thinker" large language model (LLM) agent refines gameplay strategies for an "executor" LLM agent, and a single-agent system based on refining a value function for a limited Monte Carlo Tree Search. We also experimented with rollback features to avoid performance degradation. Our results demonstrate the potential of evolutionary refinement techniques in improving AI performance in non-deterministic environments. The single-agent system achieved substantial improvements, with an average increase of 473.2 points per cycle, and with clear upward trends (correlation $ρ$=0.607) across training cycles. The LLM's understanding of the game grew as well, shown in its development of increasingly advanced strategies. Conversely, the two-agent system did not garner much improvement, highlighting the inherent limits of meta-prompting.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。