LLM-powered reasoning in agent-based modeling

2026-07-07

AI Summary

Traditional agent-based modeling (ABM) relies on static priors and struggles to capture dynamically evolving human behavior. This work proposes HALE, a hybrid simulation framework that integrates large language models (LLMs) with ABM, pioneering the use of LLMs for large-scale agent-based modeling to dynamically generate individual decision-making. By doing so, HALE substantially enhances the adaptability and scalability of simulations in complex, time-varying environments. The framework’s efficacy is empirically demonstrated through simulations of the COVID-19 pandemic in Salt Lake City, where it successfully captures the coupled dynamics between human behavioral responses and disease transmission. These results highlight HALE’s innovative potential and practical utility for modeling real-world socio-behavioral systems.

Abstract

Agent-based modeling (ABM) has the capability to model millions of individuals and their interactions, which is useful for policy making. However, ABMs have traditionally relied on static prior, which prevents the models from adapting to real-time changes. Our research provides a novel approach to addressing this information gap. Large language models (LLMs) offer new opportunities to predict human decision-making. Here, we introduce a scalable Hybrid Agent-based and Language-driven Epidemic (HALE) modeling framework that leverages LLMs to predict human decision-making in an ABM simulation. As a proof-of-concept, we use HALE to simulate COVID-19 and its effects in Salt Lake County, UT.

Problem

• agent-based modeling
• real-time adaptation
• human decision-making
• static prior
• epidemic simulation

Innovation

• Large Language Models
• Agent-based Modeling
• Human Decision-making
• Epidemic Simulation
• Hybrid Modeling