Year: 2026
Type: conference-paper
Abstract: Epidemic intervention policies are textual artefacts that human decision-makers interpret, justify, and revise through natural language, making large language models a natural candidate for epidemic policy reasoning.A naive LLM, however, lacks the epidemic dynamics needed to project intervention consequences, the quantitative surveillance signals required to assess severity, and the institutional constraints that define admissible actions.We present Epi-World, a closed-loop framework that grounds an LLM policy actor in a learned actionconditioned epidemiological world model and a tiered skill library of public-health protocols, surveillance tools, and adaptive lessons accumulated through after-action analysis.Given a candidate intervention, the world model predicts regional epidemic evolution and enables fast counterfactual rollouts that provide feedback for policy selection and refinement.Outcomes of simulated futures are distilled into reusable lessons while protocol constraints remain fixed, allowing the decision process to improve without sacrificing interpretability or controllability.We evaluate both the world model and the end-to-end framework on retrospective COVID-19 and Influenza datasets: the world model achieves the best out-of-distribution Peak-MAE among all forecasting baselines, and the closed-loop framework reduces cumulative hospitalisation by up to 59% across datasets and by an average of ∼16% across six LLM backbones, outperforming reinforcementlearning and optimal-control policy baselines.
Institutions: Emory University, Virginia Tech
Language: English
Subfield: Modeling and Simulation
Field: Mathematics
Domain: Physical Sciences
Open Access status: bronze
Awards: Collaborative Research: OAC Core: Distributed Graph Learning Cyberinfrastructure for Large-scale Spatiotemporal Prediction, A Statistics-Based Geographic Bias Quantification and Debiasing Framework for GeoAI and Foundation Models, Collaborative Research: IIBR: Innovation: Bioinformatics: Linking Chemical and Biological Space: Deep Learning and Experimentation for Property-Controlled Molecule Generation, PIPP Phase II: Community Empowering Pandemic Prediction and Prevention from Atoms to Societies (COMPASS), G6252219194, III: Medium: Towards Inclusive Recommendation Systems with Stakeholder Alignment, G8575663739
OpenAlex ID: W7220399643