当数字开始说话：LLM中的数值信号与策略行为

中文摘要 AI 辅助

基于大型语言模型（LLM）的智能体越来越多地运行在以策略互动为特征的多智能体系统（MAS）中。然而，关于不同类型的消息是否以及多大程度上影响策略博弈的结果，目前知之甚少。通过研究基于四种主流LLM的AI智能体，进行四场具有不同合作均衡的博弈，我们研究了不同类型的消息（自然语言、数值信号或随机序列）是否显著改变每场博弈中的合作水平，这也取决于智能体被分配的性格。我们观察到，结构化消息改变了大多数博弈和LLM的最终收益，但没有可预测的模式；这对AI智能体无论额外能力如何都能收敛到稳定均衡的假设提出了挑战。此外，我们观察到，智能体生成的数值消息偏离随机性，当智能体被明确指示进行通信时，这种偏离最强且最一致；然而，它们引入了额外的可解释性挑战，因为其符号分布大多与收益结构相关，并通常随着重复而变得更加集中，但总体上难以被人类解释。因此，通过受限渠道监控AI智能体的协调，应优先考虑消息级别的指纹（这些指纹可跨模型泛化），而非行为决策（这些决策不可泛化）。

英文摘要

Large language model (LLM)-based agents increasingly operate in multi-agent systems (MAS) characterised by strategic interaction. However, little is known about whether, and to what extent, different types of messages affect the outcomes of strategic games. By investigating AI agents based on four popular LLMs, playing four games with different cooperation equilibria, we study whether messages of different kinds (natural language, numerical signals, or random sequences) significantly modify the levels of cooperation in each game, also depending on the agents' assigned personalities. We observe that structured messages alter the final payoffs for most games and LLMs, but without a predictable pattern; this challenges the assumption that AI agents can converge to stable equilibria regardless of additional capabilities. Moreover, we observe that agent-generated numerical messages depart from randomness, most strongly and consistently when agents are explicitly instructed to communicate; however, they introduce an additional interpretability challenge, as their symbol distributions are mostly associated with the payoff structure and typically become more concentrated with repetition, but are overall difficult for humans to interpret. Monitoring for coordination of AI agents through restricted channels should thus prioritise message-level fingerprints, which generalise across models, over behavioural decisions, which do not.

发表机构

• Luxembourg Institute of Science and Technology（卢森堡科学技术研究院）
• University of Trento（特伦托大学）
• Teesside University（提赛德大学）
• University of Cambridge（剑桥大学）

机构由 AI 辅助整理，请以论文原文为准。