arXiv cs.MA· Tianqi Liu, Nayoung Kim, Julia Sebastien, Andrea Stevenson Won·· 6 天前AI 评分16
IMPACT：在庞贝家庭的生成式多智能体模拟中建模社会性相互依赖的移动
IMPACT: Modeling Socially Interdependent Movement in a Generative Multi-Agent Simulation of a Pompeian Household
AI 导读
研究者提出 IMPACT（Interdependent Movement Planning through Inter-Agent Constraints and Triggers）架构，用文化特定的角色与义务定义智能体活动间的依赖关系，通过社会门控里程碑规划、等待或提示解析、结构化指令发布与指令整合来引导协调。
正文 · 原文
Abstract:Simulations of archaeological sites can make interpretations of past cultural practices observable and examinable. Generative multi-agent simulations offer a bottom-up approach to modeling how people collectively moved through and used historical spaces. However, current agents designed to simulate everyday life often plan and act independently, limiting their ability to capture how movement depends on others' actions. We introduce IMPACT (Interdependent Movement Planning through Inter-Agent Constraints and Triggers), an architecture that uses culturally specific roles and obligations to define dependencies among agents' activities and guide coordination. IMPACT connects socially gated milestone planning, wait-or-prompt resolution, structured directive issuance, and directive integration. These mechanisms determine whether and when activities can begin or change as social conditions evolve, producing socially constrained and prompted movement as their primary observable outcome. We instantiate IMPACT in a five-hour simulation of a Pompeian dinner involving ten agents across interdependent roles. Analysis of five simulation runs shows how social roles, responsibilities, and status relations shape household activities and spatial practices, as reflected in patterns of co-location, asymmetric waiting, co-movement, and social directives. In a controlled ablation evaluation, thirty-seven participants rated the complete architecture's behavior as more socially coherent and believable than that of two reduced architectures. Interviews with six archaeology experts highlighted historically plausible movement patterns and the simulation's potential to support archaeological interpretation, while identifying areas requiring stronger historical grounding for future work.
| Subjects: | Multiagent Systems (cs.MA) |
| Cite as: | arXiv:2609.38113 [cs.MA] |
| (or arXiv:2609.38113v2 [cs.MA] for this version) | |
| https://doi.org/10.48550/arXiv.2609.38113 arXiv-issued DOI via DataCite |
Submission history
From: Tianqi Liu [view email]
[v1]
Tue, 29 Sep 2026 17:49:21 UTC (2,346 KB)
[v2]
Thu, 8 Oct 2026 18:49:12 UTC (2,668 KB)
来源：arXiv cs.MA · arxiv.org