Belief-Driven Multi-Agent Collaboration via Approximate Perfect Bayesian Equilibrium for Social Simulation

2026-03-25

AI Summary

This work addresses the limitations of existing large language model–based multi-agent systems, which typically rely on static interaction topologies and thus fail to capture the dynamic interplay of cooperation and competition inherent in real-world social settings, often leading to groupthink or coordination failure. To overcome this, the authors propose the BEACOF framework, which, for the first time, integrates dynamic games of incomplete information with approximate perfect Bayesian equilibrium into multi-agent social simulation. Within this framework, agents iteratively update Bayesian beliefs about peers’ capabilities and adaptively refine their collaboration strategies, enabling sequentially rational decision-making under uncertainty. This approach effectively resolves the circular dependency between collaboration-type selection and capability assessment, significantly enhancing the robustness, realism, and convergence quality of collective collaboration in domains such as legal reasoning, social interaction, and healthcare.

Technology Category

• Multiagent Systems: Mechanism Design
• Cognitive Modeling & Cognitive Systems: Social Cognition And Interaction
• Game Theory and Economic Paradigms: Coordination and Collaboration

Application Category

• Economics, Online Markets and Human Computation: Social networks and social learning
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

High-fidelity social simulation is pivotal for addressing complex Web societal challenges, yet it demands agents capable of authentically replicating the dynamic spectrum of human interaction. Current LLM-based multi-agent frameworks, however, predominantly adhere to static interaction topologies, failing to capture the fluid oscillation between cooperative knowledge synthesis and competitive critical reasoning seen in real-world scenarios. This rigidity often leads to unrealistic ``groupthink'' or unproductive deadlocks, undermining the credibility of simulations for decision support. To bridge this gap, we propose \textit{BEACOF}, a \textit{belief-driven adaptive collaboration framework} inspired by Perfect Bayesian Equilibrium (PBE). By modeling social interaction as a dynamic game of incomplete information, BEACOF rigorously addresses the circular dependency between collaboration type selection and capability estimation. Agents iteratively refine probabilistic beliefs about peer capabilities and autonomously modulate their collaboration strategy, thereby ensuring sequentially rational decisions under uncertainty. Validated across adversarial (judicial), open-ended (social) and mixed (medical) scenarios, BEACOF prevents coordination failures and fosters robust convergence toward high-quality solutions, demonstrating superior potential for reliable social simulation. Source codes and datasets are publicly released at: https://github.com/WUT-IDEA/BEACOF.

Problem

• multi-agent collaboration
• social simulation
• dynamic interaction
• groupthink
• coordination failure

Innovation

• Perfect Bayesian Equilibrium
• belief-driven collaboration
• multi-agent social simulation
• adaptive interaction topology
• incomplete information game