WhatIf: Interactive Exploration of LLM-Powered Social Simulations for Policy Reasoning

2026-04-19

AI Summary

This work proposes the first system that integrates large language model (LLM)-driven social simulation into an interactive reasoning environment to support policymaking under deep uncertainty. The system enables streaming control, real-time scalability, collaborative exploration, and multi-level interpretability by combining LLM-based agent simulation, an interactive interface, and multi-granularity behavioral visualization. Evaluated across three evacuation scenarios, it empowers expert users to uncover implicit assumptions and planning gaps, and facilitates iterative strategy comparison and refinement through traceable individual-level behaviors.

Technology Category

• Multiagent Systems: Agent-Based Simulation and Emergent Behavior
• Cognitive Modeling & Cognitive Systems: Simulating Human Behavior
• Machine Learning: Large Multimodal Models (LMMs)

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

Policymakers in domains such as emergency management, public health, and urban planning must make decisions under deep uncertainty, where outcomes depend on how large populations interpret information, coordinate, and adopt over time. Existing tools only partially support this process: tabletop exercises enable collaborative discussion but lack dynamic feedback, while computational simulations capture population dynamics but are designed for offline analysis. We present WhatIf, an interactive system that enables policymakers to steer, inspect, and compare LLM-powered social simulations in real time. Informed by a formative study in emergency preparedness planning, we derive four design requirements for interactive policy simulations: fluid steering, real-time scale, collaborative exploration, and multi-level interpretability. We developed WhatIf guided by these requirements and evaluated it with five preparedness professionals across three disaster evacuation scenarios. Our findings show that participants used the system as a space for iterative branching and comparison rather than evaluating fixed plans; reflected on tacit planning assumptions when agent behavior violated expectations; surfaced previously unrecognized planning vulnerabilities; and grounded their reasoning in inspectable agent-level cases rather than aggregate outputs alone. These findings suggest broader design implications for LLM-powered social simulation systems: designing such systems as interactive, shared reasoning environments -- rather than offline predictive tools -- can better support expert decision-making under deep uncertainty.

Problem

• policy reasoning
• deep uncertainty
• social simulations
• LLM-powered systems
• interactive exploration

Innovation

• interactive simulation
• LLM-powered social simulation
• policy reasoning
• multi-level interpretability
• real-time steering