CitySim: Modeling Urban Behaviors and City Dynamics with Large-Scale LLM-Driven Agent Simulation

2025-06-26

AI Summary

Existing urban simulation methods rely on handcrafted rules, limiting their ability to model individual intent, dynamic adaptation, and long-term planning. This paper introduces the first city-scale multi-agent simulation system deeply integrated with large language models (LLMs). We propose a “Belief–Goal–Memory” triadic cognitive architecture and a recursive value-driven mechanism to generate fine-grained daily plans, enabling tens of thousands of agents to co-evolve over extended time horizons within real-world urban geospatial contexts. Our framework is the first to endow simulated agents with explicit intent modeling, adaptive decision-making, and socially emergent behavioral evolution—overcoming expressivity bottlenecks inherent in rule-based paradigms. Experiments demonstrate significant improvements over baselines in both micro-level behavioral fidelity (e.g., path selection, activity sequencing) and macro-level urban phenomena prediction (e.g., pedestrian density, venue popularity, resident well-being), while maintaining scalability and interpretability.

Technology Category

Multiagent Systems: Agent-Based Simulation and Emergent BehaviorCognitive Modeling & Cognitive Systems: Agent ArchitecturesPlanning, Routing, and Scheduling: Planning with Language Models

Application Category

Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactionsUser Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendationSocial Networks and Social Media: Generative AI / large language models and their impact on social systems

Abstract

Modeling human behavior in urban environments is fundamental for social science, behavioral studies, and urban planning. Prior work often rely on rigid, hand-crafted rules, limiting their ability to simulate nuanced intentions, plans, and adaptive behaviors. Addressing these challenges, we envision an urban simulator (CitySim), capitalizing on breakthroughs in human-level intelligence exhibited by large language models. In CitySim, agents generate realistic daily schedules using a recursive value-driven approach that balances mandatory activities, personal habits, and situational factors. To enable long-term, lifelike simulations, we endow agents with beliefs, long-term goals, and spatial memory for navigation. CitySim exhibits closer alignment with real humans than prior work, both at micro and macro levels. Additionally, we conduct insightful experiments by modeling tens of thousands of agents and evaluating their collective behaviors under various real-world scenarios, including estimating crowd density, predicting place popularity, and assessing well-being. Our results highlight CitySim as a scalable, flexible testbed for understanding and forecasting urban phenomena.

Problem

• Modeling nuanced human behaviors in urban environments
• Simulating adaptive agent actions with long-term goals
• Forecasting collective urban phenomena at large scales

Innovation

• LLM-driven agent simulation for urban behaviors
• Recursive value-driven daily schedule generation
• Beliefs, goals, spatial memory for lifelike agents