🤖 AI Summary
Quantifying the impact of human behavior on disease transmission during pandemics remains highly challenging. This work proposes the Epi-LLM framework, which for the first time integrates large language models (LLMs) into an agent-based SEIR epidemic model, leveraging data from pandemic-related behavioral game experiments and generalized linear models to simulate agents’ dynamic reasoning and adaptive behaviors within contact networks. The study reveals that architectural variance among LLMs significantly affects the validity of behavioral simulations and demonstrates the necessity of explicitly parameterizing attitudes to capture cultural differences. Experimental results show that all four LLM architectures effectively reduce infection peaks, achieving quarantine compliance rates of 58–65%. Perceived health severity emerges as the strongest behavioral predictor, and the model’s pseudo-R² aligns closely with empirical findings from human experiments.
📝 Abstract
Human behaviour during epidemics affects infectious disease dynamics, but quantifying this remains deeply challenging. Here we introduce the Epi-LLM framework: a novel integration of agent-based modelling, real-life epigames, and large language models (LLMs) in which a synthetic society of agents reasons and adapts dynamically over an outbreak contact network. Comparing synthetic agent behaviour against a no-intervention SEIR baseline and human participant data from the AUIB epigame study, we find that LLM agents across four different architectures reduced peak active infections, with quarantine compliance peaking at 58-65% on day six of the 15-day simulation. A binomial generalised linear model showed that perceived health severity was the strongest predictor of quarantine behaviour ($β= 0.33, p = 0.002$), yielding a pseudo-$R^2$ of 0.055, comparable to the 0.072 observed in the human trial. LLM architecture is a key determinant of epidemic dynamics: low-variance architectures offer greater internal validity for testing behavioural rules, while high-variance models may better represent real-world decision-making. Geographic labels alone do not induce culturally differentiated behaviour; explicit attitudinal parameterisation is required. This proof-of-principle work lays the groundwork for deploying the Epi-LLM framework as a scalable, risk-free simulation environment for pandemic preparedness research.