IMPACT: Modeling Socially Interdependent Movement in a Generative Multi-Agent Simulation of a Pompeian Household

2026-09-29

AI Summary

This study addresses the limitation of existing simulation agents acting independently, which fails to capture socially interdependent movement. To overcome this, we propose the IMPACT architecture, introducing a collaborative framework that integrates social gating milestones, wait/prompt mechanisms, and structured instruction issuance. By defining inter-agent dependencies through cultural roles, the method coordinates activities via generative multi-agent simulation, incorporating role-based dependency modeling and dynamic constraint resolution algorithms. Experiments conducted in a Pompeii dinner scenario demonstrate that the proposed architecture generates movement patterns with significant social coherence. Ablation studies confirm its superiority over simplified baselines, and the approach has been validated by archaeological experts.

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.

Problem

• multi-agent simulation
• interdependent movement
• social coordination
• archaeological simulation
• generative agents

Innovation

• Generative Multi-Agent Simulation
• Interdependent Movement Planning
• Socially Gated Milestone Planning
• Agent Coordination
• Archaeological Simulation