用开源游戏测试大模型编程能力，发现其能自动生成合作与欺骗策略。
Evaluating LLMs in Open-Source Games
- 让大模型在开源游戏中编写程序参与博弈，利用代码透明性探索新策略
- 模型生成了追求收益最大化、合作及欺骗等多类策略并展现适应性演化
- 适合研究多智能体协作机制或评估大模型行为可预测性的研究人员
大型语言模型（LLMs）的编程能力使其能够参与开源游戏——一种玩家以计算机程序替代行动的游戏理论场景。此类程序具有可解释性、智能体间透明性及形式可验证性，并支持程序均衡，即利用代码透明性实现的解，这类解在传统标准形式博弈中无法获得。我们评估了主流开源与闭源大模型在预测和分类程序策略方面的能力，并分析了大模型智能体在双人及演化设置下所达成的近似程序均衡特征。研究发现，模型会自发生成追求收益最大化、合作及欺骗等策略，揭示其在重复游戏中的机制演化过程，并比较其相对进化适应度。结果表明，开源游戏为研究和引导多智能体困境中的合作策略涌现提供了可行环境。
原文摘要 · Abstract (English)
Large Language Models' (LLMs) programming capabilities enable their participation in open-source games: a game-theoretic setting in which players submit computer programs in lieu of actions. These programs offer numerous advantages, including interpretability, inter-agent transparency, and formal verifiability; additionally, they enable program equilibria, solutions that leverage the transparency of code and are inaccessible within normal-form settings. We evaluate the capabilities of leading open- and closed-weight LLMs to predict and classify program strategies and evaluate features of the approximate program equilibria reached by LLM agents in dyadic and evolutionary settings. We identify the emergence of payoff-maximizing, cooperative, and deceptive strategies, characterize the adaptation of mechanisms within these programs over repeated open-source games, and analyze their comparative evolutionary fitness. We find that open-source games serve as a viable environment to study and steer the emergence of cooperative strategy in multi-agent dilemmas.
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。