CS Paper Radar

EpiWorld: Grounding LLM Policy Agents in Epidemiological World Models

Zeeshan Memon, Yiqi Su, Kai Shu, Naren Ramakrishnan, Liang Zhao

cs.CL ·submitted 2026-10-02 03:16 UTC · v1
top venue

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 EpiWorld, 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 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 reinforcement-learning and optimal-control policy baselines.

Comment
Accepted to Findings of EMNLP 2026. 22 pages

Closest papers

Nearest neighbours by meaning across the whole index.

Graph World Models for Constrained Epidemic Policy Planning

cs.LG cs.AI ·2026-09-28 ·Yiqi Su, Rashed Shelim, Lingyi Wang, Walid Saad, Naren Ramakrishnan
similarity 0.72
Abstract

Epidemic policy planning often requires coordination between geographical regions, taking into account mobility-driven spillovers and how to make use of limited resources. Existing methods either lack action-conditioned models of coupled dynamics or cannot guarantee per-period feasibility. We present EpiMind, a graph world model framework for constrained epidemic policy planning across regions. A graph-factored recurrent state-space model generates joint policy-conditioned rollouts from regional latent beliefs, while graph-temporal ADMM optimizes regional interventions, enforces shared-resource feasibility through projection, and evaluates temporal specifications under the learned model. EpiMind reduces admission RMSE by 29% relative to graph-free dynamics modeling, plans within 1-5% of the best feasible constant policy with guaranteed shared-budget feasibility, and outperforms all deployable baselines across three resource budgets in real-context evaluation. These results demonstrate that graph-structured policy imagination with explicit constrained coordination supports effective epidemic interventions from learned dynamics.

0.25

Integrating adaptive human behavior into epidemic models with large language models

cs.AI physics.soc-ph ·2026-08-30 ·Yicheng Mao, Haoyang Li, Rob Deardon, Hongru Du
similarity 0.69
Abstract

Infectious disease transmission is shaped by patterns of human interaction, which adapt as epidemic conditions change. Capturing these context-dependent behaviors remains a fundamental challenge for epidemic models. Here, we recast this challenge by using large language models (LLMs) to represent adaptive human behavior within mechanistic epidemic models. We operationalize this idea through Generative Adaptive Behavioral Layer for Epidemics (GABLE), which adapts LLMs to infer behavioral responses to epidemic and policy conditions and translates them into age-structured contact matrices coupled to a mechanistic epidemic model. Applied to COVID-19 in France, GABLE reproduced responses in population mixing and age-specific contact structures that remained epidemiologically informative. In short-term forecasting, LLM-generated contact matrices outperformed mobility-driven matrices derived from real-world mobility data, with the largest gains at longer horizons. GABLE also extends beyond forecasting to prospective policy evaluation by projecting behavioral and epidemic responses to candidate interventions before implementation. When supplied with subsequently implemented policies, GABLE reproduced epidemic trajectories and generated distinct responses to alternative policy timing and composition. By leveraging LLMs as a flexible behavioral layer, GABLE provides a framework for coupling context-sensitive behavioral generation with epidemic dynamics.

0.22

EPI-SST: Renewal-Structured State-Space Transformers for Epidemic Forecasting

cs.LG ·2026-11-16 ·Yiyun Chen, Zewen Liu, Shengpu Tang, Max S. Y. Lau, Wei Jin
similarity 0.65
Abstract · 13 pages, 4 figures, 12 tables

Epidemic forecasting supports public health decisions ranging from capacity planning to intervention timing, for which both point accuracy and calibrated uncertainty are essential. Although neural models have advanced point forecasting, producing reliable probabilistic predictions over multiple future horizons remains challenging. Renewal models encode epidemic propagation, while neural forecasters learn flexible multi-horizon representations; integrating these strengths within a probabilistic future rollout remains underexplored. We propose EPI-SST (Epidemic State Space Transformer), a probabilistic forecaster centered on an evidence-driven renewal transition. At each horizon, neural evidence produces a bounded update to an internal multiplier and advances a finite-memory renewal state that generates the future mean trajectory. A separate heteroscedastic Gaussian head controls predictive scale using a validation-selected log-variance penalty. Across four epidemic surveillance benchmarks, EPI-SST achieves the lowest pooled RMSE on every dataset, reducing RMSE by up to 14.2 percent relative to the strongest baseline. It also achieves the lowest continuous ranked probability score (CRPS) at horizons h = 4, 5, 6 on every dataset, with reductions of up to 43 percent relative to the strongest probabilistic baseline. During the abnormal 2019-2020 influenza season, EPI-SST reduces CRPS by 16.9 percent and calibration error by 78.9 percent; a three-seed replication preserves its distributional lead. Ablation studies and frozen-path interventions show that the renewal transition, attention conditioning, horizon coupling, and uncertainty-scale control make complementary contributions to predictive performance.

0.35

Universal Infectious Disease Forecasting with Mantis 2

cs.AI q-bio.PE ·2026-11-12 ·Carson Dudley, Reiden Magdaleno, Yutong Bi, Anthony Valencia, Xiaofeng Liu +1 more
similarity 0.65
Abstract

Public health agencies need to anticipate how outbreaks will unfold, when they will peak, and how long demand on health services will remain elevated. We previously developed a foundation model, Mantis, which was trained entirely on simulated outbreaks---enabling us to generate the diversity and scale of training data needed to support forecasting across diseases. However, our previous work left unresolved when simulation improves predictions and how it should complement observational data. Here, we varied the balance of simulated and observed training data while holding forecasting tasks fixed and varying the available surveillance history. We found that simulation-rich training was particularly valuable with limited local history, and its relative advantage diminished as more observations became available. We also found that simulation-based training reduced exceptionally large forecast errors, suggesting that mechanistic knowledge can help compensate for limited local evidence. Building on these findings, we developed Mantis 2, which combines simulated and observational training to jointly forecast up to six variables sixteen weeks ahead. Across 30 forecasting tasks, Mantis 2 achieved the best aggregate probabilistic accuracy over short, medium, and long horizons among the machine learning, statistical, and baseline models evaluated. It missed roughly 30\% fewer outbreak peaks than the strongest comparison model, with successfully anticipated peaks identified more than six weeks in advance on average. We publicly release Mantis 2, enabling public health agencies and researchers to forecast disease outcomes without developing a separate model for each disease or location.

0.26

Driving Epidemic Models with AI Agents: the Epydemix Agent Framework

cs.AI cs.CY ·2026-09-23 ·Nicolò Gozzi, Ciro Cattuto, Alessandro Vespignani
code/data similarity 0.64
Abstract

Artificial Intelligence agents based on large language models provide convenient natural language interfaces to scientific software, but reliability is not automatic. Here we introduce the Epydemix Agent Framework, an additive layer over Epydemix, an open-source Python library for stochastic compartmental epidemic modeling. The framework extends the library with four capabilities to facilitate interaction with an AI agent: discovery of available models and parameters, preventive validation of a declarative scenario specification, execution through tested library code, and inspectability of results. These capabilities let an agent handle the entire modeling process, from the natural-language description of the scenario to quantitative results, figures, and interpretation of findings without writing custom code. Each step reads input files and saves results in a separate output bundle, making the process auditable and reproducible. First, we show the end-to-end workflow with a case study comparing vaccination strategies for a novel respiratory virus. Second, we assessed the framework across 50 agent sessions and five modeling tasks by comparing the agent use of the framework against the direct use of the Python interface. The framework reduced turns, output tokens, and cost on most tasks, unless it trades resources for per-point reproducibility.

0.26

RISE: Adaptive Imagination for World Action Models

cs.CV ·2026-08-20 ·Hongbo Lu, Liang Yao, Chenghao He, Hao Han, Fan Liu +3 more
highly cited similarity 0.64
Abstract

World Action Models (WAMs) improve planning by incorporating future world evolution into action generation, yet existing methods allocate a fixed imagination budget to every scene. We propose RISE (\textbf{R}efining \textbf{I}magination through \textbf{SE}lective Rollout), a system-level adaptive imagination framework that makes sequential \textsc{Roll}/\textsc{Stop} decisions according to the expected planning benefit of continued rollout. At each step, a Latent Evaluator estimates the risk revealed by the current prefix and how much planning could improve if imagination continues, while a Rollout Gate weighs this expected benefit against additional computation cost. Since factual driving logs expose only one realized future, we further construct \textbf{CounterDrive}, a counterfactual dataset with diverse outcomes and risk levels, to enrich future dynamics and provide localized risk supervision. Each retained sample undergoes expert verification and annotation of trajectory validity, incident onset, and causal category, providing a reusable resource for safety-critical world-modeling research. Experiments on NAVSIM and nuScenes show that RISE achieves the best overall planning performance while reducing unnecessary rollout, with additional transfer results supporting its plug-in generality across WAM architectures.

0.36

From Urban Mobility to Epidemic Dynamics: A Mixture-of-Experts Framework with Preference Alignment for Policy Scenario Simulation

cs.CY ·2026-08-20 ·Yun Ye, Arsalan Dezhkam, Junyuan Liu, Xinglei Wang, Tao Cheng
similarity 0.63
Abstract

Non-pharmaceutical interventions (NPIs) alter epidemic risk through behavioral reallocations, not simply aggregate mobility reductions. Scenario-based NPI analysis therefore requires a behavioral layer that translates alternative policy calendars into plausible activity and mobility trajectories before downstream outcomes are simulated. We introduce UrbanShare-MoE-PA, a data-driven agent-level framework that maps factual and alternative NPI calendars to daily time-allocation trajectories and propagates them through a calibrated behavior-driven SEIR simulator. The behavioral engine decomposes each agent-day into travel share, POI-category allocation conditional on staying, and travel-mode allocation conditional on traveling. It combines a structured UrbanShare baseline, mixture-of-experts heads for heterogeneous POI and mode responses, and phase-aware preference alignment for calendar-conditioned rollouts. Using data from 911 agents in Singapore observed from March to August 2020, we evaluate factual reconstruction, four alternative lockdown calendars, and epidemic-activity trade-offs. UrbanShare-MoE improves POI reconstruction over the baseline, while UrbanShare-MoE-PA achieves the lowest travel-mode errors and the clearest alternative-calendar trajectories. In the calibrated SEIR simulation, early lockdown lowers infectious burden, late lockdown increases it, and short lockdown preserves the highest weighted activity with only a modest increase in epidemic burden relative to the original policy. These results show that epidemic-activity conclusions depend on how policy calendars are translated into behavior, and that agent-level mobility-share modeling provides an interpretable bridge between policy timing, behavior, and downstream simulation.

0.34

EpiGuard: An Autonomous Infectious-Disease Early Warning Agent Platform

cs.AI ·2026-10-14 ·Tonglong Wei, Zheng Yuan, Lingshuai Lin, Yuhui Yao, Xing Xie +1 more
similarity 0.62
Abstract

Infectious disease outbreaks often spread rapidly before signs of widespread transmission become apparent, leaving a big challenge for rapid response and effective intervention. However, current infectious-disease early-warning systems face three major challenges: fragmented epidemiological evidence across diverse surveillance sources, highly uneven data availability across diseases and geographic regions, and delayed warning procedure caused by manual data collection, analysis, and reporting. To address these challenges, we present EpiGuard, an autonomous, continually self-improving agentic platform for timely and reliable early warning of more than 50 infectious diseases across different regions. EpiGuard proactively monitors and captures newly released official surveillance records, using the latest available evidence for risk assessments. It dynamically adjusts its analytical strategy across three evidence availability regimes, enabling reliable risk assessment under diverse data conditions, ranging from the absence of historical observations to sparse and abundant evidence. Moreover, EpiGuard learns from the successful execution process by preserving effective data-acquisition and analytical procedures as reusable standard operating procedures, enabling it to progressively generalize to previously unseen diseases and regions. We use officially released infectious-disease surveillance data from the China CDC and the United States CDC to evaluate EpiGuard's effectiveness. Experimental results demonstrate that EpiGuard outperforms state-of-the-art forecasting systems, improving forecasting performance by 9.68% and 14.04% on the China CDC and U.S. CDC datasets, respectively.

0.20