Large Language Models Empowered Agent-based Modeling and Simulation: A Survey and Perspectives
Chen Gao Xiaochong Lan Nian Li Yuan Yuan Jingtao Ding Zhilun Zhou Fengli Xu Yong Li Tsinghua University, Beijing, China {chgao96, fenglixu,
Abstract
Agent-based modeling and simulation has evolved as a powerful tool for modeling complex systems, offering insights into emergent behaviors and interactions among diverse agents. Integrating large language models into agent-based modeling and simulation presents a promising avenue for enhancing simulation capabilities. This paper surveys the landscape of utilizing large language models in agent-based modeling and simulation, examining their challenges and promising future directions. In this survey, since this is an interdisciplinary field, we first introduce the background of agent-based modeling and simulation and large language model-empowered agents. We then discuss the motivation for applying large language models to agent-based simulation and systematically analyze the challenges in environment perception, human alignment, action generation, and evaluation. Most importantly, we provide a comprehensive overview of the recent works of large language model-empowered agent-based modeling and simulation in multiple scenarios, which can be divided into four domains: cyber, physical, social, and hybrid, covering simulation of both real-world and virtual environments. Finally, since t
中文速览
把大语言模型（LLM）引入基于智能体的建模与仿真（Agent-Based Modeling and Simulation, ABMS）是一个充满潜力却缺乏系统梳理的新兴方向，这篇综述正是为填补这一空白而写。作者从感知、推理、决策、适应性和异质性等维度分析了LLM如何克服传统仿真方法在智能体能力上的瓶颈，使智能体具备更接近人类的自主性、社会交互能力和目标导向行为。在此基础上，论文将现有工作系统归纳为网络空间、物理世界、社会领域和混合场景四大类，覆盖从交通、生态、经济市场到流行病传播的广泛应用，详细介绍了各类场景下仿真环境与LLM驱动智能体的设计方法。最后，作者指出规模化仿真效率、开放平台建设、鲁棒性以及伦理风险等尚待解决的关键问题，为这一快速演进的跨学科领域勾勒出未来研究路线图。这项工作的重要性在于：它是首个系统整合LLM与ABMS两大领域的综述，有助于研究者快速把握现状、找准切入点，推动更真实、更智能的复杂系统仿真研究。
原文 arXiv:2312.11970;中英对照 + 大白话阅读 https://aha.fim.ai/paper/2312.11970v1