用心理分析框架让大模型智能体更像真人互动。
Games Agents Play: Towards Transactional Analysis in LLM-based Multi-Agent Systems
- 将心理分析中的父母/成人/儿童三状态融入智能体认知架构
- 在愚蠢游戏模拟中生成更具情境感知的深度交互
- 适合社会心理学、教育辅导等需要真实人际动态的场景
多智能体系统（MAS）日益用于模拟社会互动，但现有框架普遍缺乏人类行为背后的心理复杂性。本文提出Trans-ACT（交易分析认知工具包），将交易分析（TA）原理融入MAS，使智能体具备真实的心理动态。Trans-ACT在智能体认知架构中整合了父母、成人、儿童三种自我状态，每种状态会调用与情境相关的记忆，并据此影响对新情境的回应。最终输出由智能体潜在的生活脚本决定。实验模拟‘愚蠢游戏’场景显示，基于认知与TA原则的智能体产生了更深刻且情境敏感的交互。未来研究可拓展至冲突调解、教育支持及高级社会心理学研究等领域。
原文摘要 · Abstract (English)
Multi-Agent Systems (MAS) are increasingly used to simulate social interactions, but most of the frameworks miss the underlying cognitive complexity of human behavior. In this paper, we introduce Trans-ACT (Transactional Analysis Cognitive Toolkit), an approach embedding Transactional Analysis (TA) principles into MAS to generate agents with realistic psychological dynamics. Trans-ACT integrates the Parent, Adult, and Child ego states into an agent's cognitive architecture. Each ego state retrieves context-specific memories and uses them to shape response to new situations. The final answer is chosen according to the underlying life script of the agent. Our experimental simulation, which reproduces the Stupid game scenario, demonstrates that agents grounded in cognitive and TA principles produce deeper and context-aware interactions. Looking ahead, our research opens a new way for a variety of applications, including conflict resolution, educational support, and advanced social psychology studies.
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。