# Epidemiological World Model

Epidemic interventions are textual policies, yet a plain LLM lacks the epidemic dynamics to project their consequences, the surveillance signals to judge severity, and the institutional constraints on admissible actions. EpiWorld grounds an LLM policy agent in an action-conditioned epidemiological world model, screens candidate interventions through counterfactual rollouts, and distils simulated outcomes into reusable lessons. On retrospective COVID-19 and influenza data, the full configuration cuts 8-week cumulative hospitalisations by 59.3% versus the reference historical policy on real US Omicron, and by about 16% on average across six LLM backbones.

Epidemic interventions are textual policies, yet a plain LLM lacks the epidemic dynamics to project their consequences, the surveillance signals to judge severity, and the institutional constraints on admissible actions. EpiWorld grounds an LLM policy agent in an action-conditioned epidemiological world model, screens candidate interventions through counterfactual rollouts, and distils simulated outcomes into reusable lessons. On retrospective COVID-19 and influenza data, the full configuration cuts 8-week cumulative hospitalisations by 59.3% versus the reference historical policy on real US Omicron, and by about 16% on average across six LLM backbones.

EpiWorld is a closed-loop framework that grounds an LLM policy actor in a learned action-conditioned epidemiological world model and a tiered skill library of public-health protocols, surveillance tools, and lessons accumulated through after-action analysis; the world model predicts regional epidemic evolution and enables fast counterfactual rollouts that feed policy selection and refinement, and on retrospective COVID-19 and Influenza datasets it achieves the best out-of-distribution Peak-MAE among forecasting baselines while the closed-loop framework reduces cumulative hospitalisation by up to 59% across datasets and by an average of about 16% across six LLM backbones, outperforming reinforcement-learning and optimal-control policy baselines.

EpiWorld is a closed-loop framework that grounds an LLM policy actor in a learned action-conditioned epidemiological world model and a tiered skill library of public-health protocols, surveillance tools, and lessons accumulated through after-action analysis; the world model predicts regional epidemic evolution and enables fast counterfactual rollouts that feed policy selection and refinement, and on retrospective COVID-19 and Influenza datasets it achieves the best out-of-distribution Peak-MAE among forecasting baselines while the closed-loop framework reduces cumulative hospitalisation by up to 59% across datasets and by an average of about 16% across six LLM backbones, outperforming reinforcement-learning and optimal-control policy baselines.