LLM-Driven Personalities for Decision Making in Emergency Simulations

2026-06-29

AI Summary

This work proposes an intelligent decision-making framework that integrates large language models (LLMs) with the OCEAN five-factor personality model to enhance behavioral realism and diversity in virtual human agents during emergency evacuation simulations. By designing personality-guided prompt templates, the approach enables LLMs to generate individualized, context-aware decisions that reflect distinct personality traits. This represents the first application of combining LLMs with formal personality models in emergency evacuation scenarios, overcoming the limitations of traditional rule-based systems in capturing behavioral heterogeneity and expressiveness. Experimental results demonstrate that varying personality configurations significantly influence individual route choices and emergent crowd dynamics, thereby substantially improving the diversity and realism of simulated populations.

Abstract

For virtual humans to appear believable, they must exhibit agency and spatial awareness while interacting with their environment in ways that reflect competence and intelligence. At the core of these capabilities lies effective decision-making, which strongly shapes agent behavior. With the rapid advancement of artificial intelligence, Large Language Models (LLMs) have increasingly been explored as a mechanism to support such decision-making processes. In this work, we investigate the use of LLMs to drive decision-making in virtual humans within a simulated evacuation scenario, incorporating OCEAN personality traits into agent representations. Our goal is to evaluate how personality, expressed through language-based prompts, influences both individual behaviors and collective simulation outcomes. Our results demonstrate that LLM-driven personality profiles significantly impact agents' decisions, leading to distinct behavioral patterns across different traits. These findings suggest that heterogeneous crowds composed of LLM-guided agents can enhance the realism and variability of simulated environments, offering a flexible alternative to traditional rule-based approaches.

Problem

• virtual humans
• emergency simulation
• decision-making
• personality traits
• Large Language Models

Innovation

• Large Language Models
• OCEAN personality traits
• virtual humans
• emergency simulation
• agent decision-making