EconSimulacra: A Digital Twin Platform of Socio-Economic Systems Powered by LLM Agents

2026-06-25

AI Summary

Existing social simulators struggle to capture the dynamic coupling among economic activity, mobility, and social networks. To address this limitation, this work proposes a large language model–based multi-agent digital twin platform that innovatively incorporates shared internal states—such as stress levels—to jointly model individual memory and cross-domain states. This unified framework enables coherent decision-making grounded in agents’ cross-domain experiences and facilitates the co-evolution of consumption behaviors, population mobility, and social interactions. The platform successfully reproduces the nonlinear relationship between online social attention and offline location popularity, demonstrating its effectiveness in generating realistic, cross-domain socio-behavioral dynamics.

Abstract

Real-world social behavior emerges from tightly coupled domains: economic conditions shape mobility and social interactions, while online attention and offline activity feed back into local popularity and consumer behavior. Capturing these feedback loops requires artificial societies in which agents carry experiences from one domain into decisions in another. Large language models (LLMs) provide a promising foundation for such societies. However, existing LLM-based simulators typically model domains in isolation or merely place them side by side. To enable such cross-domain interactions, we present EconSimulacra, a multi-agent social simulator that couples consumer economy, mobility, and social networks through a shared internal-state mechanism. In EconSimulacra, experiences accumulated across different domains are stored in memory and transformed into shared internal states (i.e., stress level) connecting heterogeneous domains through individual decision making. This design allows agents to reconcile competing demands arising from multiple domains and generate coherent cross-domain behaviors. As a case study, we show that the shared internal state mechanisms reproduce a nonlinear relationship between online social attention and offline local popularity, illustrating how realistic cross-domain dynamics can emerge within a unified artificial society.

Problem

• cross-domain interactions
• socio-economic systems
• feedback loops
• artificial societies
• LLM agents

Innovation

• digital twin
• LLM agents
• cross-domain interaction
• shared internal state
• multi-agent simulation