Toward World Models for Epidemiology
2026-04-10 • Machine Learning
Machine Learning
AI summaryⓘ
The authors explain that world models, which help predict how things change over time, are useful for understanding and managing epidemics but haven't been fully used in this field. They show that epidemics are tricky to predict because the real situation is hidden, the data we see is imperfect and affected by policies, and people’s behavior changes how interventions work. The authors propose a way to think about epidemics as systems where what we observe is incomplete and actions influence future outcomes. They give three examples demonstrating why having explicit models of how the epidemic unfolds is important for making informed decisions.
world modelslatent dynamicscomputational epidemiologypartial observabilitypolicy-dependent surveillanceinterventionsadaptive behaviordynamical systemscounterfactual analysis
Authors
Zeeshan Memon, Yiqi Su, Christo Kurisummoottil Thomas, Walid Saad, Liang Zhao, Naren Ramakrishnan
Abstract
World models have emerged as a unifying paradigm for learning latent dynamics, simulating counterfactual futures, and supporting planning under uncertainty. In this paper, we argue that computational epidemiology is a natural and underdeveloped setting for world models. This is because epidemic decision-making requires reasoning about latent disease burden, imperfect and policy-dependent surveillance signals, and intervention effects are mediated by adaptive human behavior. We introduce a conceptual framework for epidemiological world models, formulating epidemics as controlled, partially observed dynamical systems in which (i) the true epidemic state is latent, (ii) observations are noisy and endogenous to policy, and (iii) interventions act as sequential actions whose effects propagate through behavioral and social feedback. We present three case studies that illustrate why explicit world modeling is necessary for policy-relevant reasoning: strategic misreporting in behavioral surveillance, systematic delays in time-lagged signals such as hospitalizations and deaths, and counterfactual intervention analysis where identical histories diverge under alternative action sequences.