Physics of Agents: Statistical Mechanics Predicts Collective Behavior of AI Agents

2026-08-17

AI Summary

This study addresses the challenges of predicting collective behavior and polarization bias in multi-agent interactions by proposing a statistical mechanics-based framework for modeling communities of ten-thousand-scale language model agents. Through formalized opinion dynamics and large-scale simulations, we reveal three evolutionary states—apathy, polarization, and consensus—driven by social temperature and attraction mechanisms. Experimental results demonstrate that the proposed model significantly outperforms baselines in predicting both individual trajectories and population-level distributions. Furthermore, it elucidates the underlying mechanisms behind improved accuracy on objective questions and right-leaning bias in subjective responses. This work establishes a novel paradigm for understanding the social dynamics of generative agent societies, bridging computational social science and large language model simulation.

Abstract

AI agents increasingly operate as part of interacting systems rather than in isolation. As agents exchange information and jointly make decisions, their interactions can improve collective reasoning but may also produce herding, polarization, or amplify shared biases. Understanding and predicting these collective dynamics is therefore important for designing effective and aligned multi-agent systems. Here, we study over 10,000 communities of language-model agents that repeatedly exchange messages and revise their opinions across objective mathematics questions and subjective political statements. Despite substantial diversity in possible behavior, the individual and group dynamics can be represented by three characteristic regimes: indifference, polarization, and consensus. AI agents start indifferent and build conviction as they interact. On objective questions, communication improves collective accuracy, while on subjective questions it often drifts group opinions toward the right in the political spectrum. We explain these observations with a statistical-mechanics formalism in which agents stochastically favor lower social pressure. Given only initial opinions, our model predicts individual trajectories, outperforms all standard baselines, generalizes to unseen community graphs, and reproduces the observed group archetype distributions. Our fitted model parameters reveal the mechanics underlying our key observations: i) communities operate below the critical social temperature, which explains conviction buildup; ii) attractive ties outweigh repulsive ones, which favors consensus; and iii) agents holding the correct answer exert the strongest pull, which drives truth-seeking. Overall, our results demonstrate that collective behavior of AI agents, like that of other complex systems, follows compact and predictive dynamical laws.

Problem

• Collective Behavior
• Multi-Agent Systems
• AI Agents
• Group Dynamics
• Social Interaction

Innovation

• Statistical Mechanics
• Collective Behavior
• Multi-Agent Systems
• Social Temperature
• Dynamical Laws