arXiv cs.MA· Guo Chen, Huaijin Xie, Ziwen Li, Reed Li, Yu Lu, Haibo Shi, Bingbing Xu, Junjie Huang·· 4 天前AI 评分27
当极端者进入 LLM 社会：EXAM 框架如何研究多智能体系统中的极端行为
When Extremists Enter LLM Societies: Investigating Extreme Behavior in Multi-Agent Systems
AI 导读
研究者提出 EXAM（EXtremist Agent Modeling）框架，可为 LLM 多智能体系统配置极端智能体，并精确控制其人口比例、空间分布、交互协议与测量接口。在国会选举、金融危机、疫情舆论和迭代囚徒困境四个场景中，规模小但持续存在的尾部群体均带来可测量的集体偏移，且该偏移依赖骨干模型与具体场景。
正文 · 原文
Abstract:Large language model (LLM) agents increasingly serve as computational actors in multi-agent social simulations. While existing work emphasizes representative personas and average-case behavior, social science shows that persistent tail actors can disproportionately shape collective dynamics. We study this regularity in controlled LLM-based multi-agent systems (LLM-MAS). In these systems, such actors cannot be reduced to fixed scalar states: prompts, natural-language histories, and backbone generation mediate their behavior and influence. We first analyzed three properties of extremists in sociology and model the LLM extremist based on LLM-native perspective. We introduce EXAM, an EXtremist Agent Modeling framework that capable of configuring extremist agents for specific tasks and precisely controlling their population proportion, spatial distribution, interaction protocols, and measurement interfaces. Across four scenarios: congressional election, financial crisis, epidemic public opinion, and iterated prisoner's dilemma, we observe that small persistent tail groups produce measurable, backbone- and scenario-dependent collective shifts. EXAM shows how this social-science regularity manifests through language-mediated mechanisms specific to LLM-based social simulation.
| Subjects: | Multiagent Systems (cs.MA) |
| Cite as: | arXiv:2610.19863 [cs.MA] |
| (or arXiv:2610.19863v1 [cs.MA] for this version) | |
| https://doi.org/10.48550/arXiv.2610.19863 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Guo Chen [view email]
[v1]
Wed, 14 Oct 2026 09:27:13 UTC (2,932 KB)
来源：arXiv cs.MA · arxiv.org