CityReal: Human-Aligned Urban Behavior and City Dynamics Simulation with Large-Scale LLM Agents
CityReal is a framework for simulating human behavior in cities using large language models. Unlike previous methods that rely on few-shot prompting, CityReal models agents as intention-driven decision makers that adapt over time based on experience and constraints. The framework includes textual adapters to improve population-level realism by aligning agent decisions with observed statistics. Experiments show improved alignment with real-world human behavior at both micro an
CityReal is a framework for simulating human behavior in cities using large language models. Unlike previous methods that rely on few-shot prompting, CityReal models agents as intention-driven decision makers that adapt over time based on experience and constraints. The framework includes textual adapters to improve population-level realism by aligning agent decisions with observed statistics. Experiments show improved alignment with real-world human behavior at both micro and macro levels.
---
Why it matters: This matters because it provides a more realistic way to simulate urban dynamics, which can inform transportation policy and social science research. By modeling agents as intention-driven decision makers, CityReal can help analyze crowd density, place popularity, and mobility flows under different urban scenarios.
Source: https://arxiv.org/abs/2608.16897
This article was originally published at: https://arxiv.org/abs/2608.16897