Evolution of Cooperation in LLM-Agent Societies: A Preliminary Study Using Different Punishment Strategies

2025-04-28

AI Summary

This study investigates whether cooperative behavior can sustainably evolve in a society of large language model (LLM) agents situated in an embodied, natural-language-driven simulation, thereby testing the applicability of the Boyd–Richerson model of cultural evolution. We introduce the first multi-agent simulation framework grounded in the “scrounger’s dilemma” game, integrating explicit punishment mechanisms and natural-language-based social reasoning, and propose a pairwise imitation strategy-updating rule to approximate human-like norm acquisition. Our results demonstrate: (1) canonical dynamics of cooperation evolution—such as norm emergence, stabilization, and collapse—are faithfully reproduced in LLM agents; (2) explicit punishment robustly induces and sustains cooperative norms across diverse initial strategy distributions, significantly elevating population-level cooperation rates; and (3) the framework establishes an interpretable, experimentally controllable testbed for empirically evaluating theories of social norm evolution using language models.

Technology Category

• Multiagent Systems: Agent-Based Simulation and Emergent Behavior
• Natural Language Processing: (Large) Language Models
• Game Theory and Economic Paradigms: Cooperative Game Theory

Application Category

• Social Networks and Social Media: Generative AI / large language models and their impact on social systems
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation

Abstract

The evolution of cooperation has been extensively studied using abstract mathematical models and simulations. Recent advances in Large Language Models (LLM) and the rise of LLM agents have demonstrated their ability to perform social reasoning, thus providing an opportunity to test the emergence of norms in more realistic agent-based simulations with human-like reasoning using natural language. In this research, we investigate whether the cooperation dynamics presented in Boyd and Richerson's model persist in a more realistic simulation of the diner's dilemma using LLM agents compared to the abstract mathematical nature in the work of Boyd and Richerson. Our findings indicate that agents follow the strategies defined in the Boyd and Richerson model, and explicit punishment mechanisms drive norm emergence, reinforcing cooperative behaviour even when the agent strategy configuration varies. Our results suggest that LLM-based Multi-Agent System simulations, in fact, can replicate the evolution of cooperation predicted by the traditional mathematical models. Moreover, our simulations extend beyond the mathematical models by integrating natural language-driven reasoning and a pairwise imitation method for strategy adoption, making them a more realistic testbed for cooperative behaviour in MASs.

Problem

• Study cooperation evolution in LLM-agent societies using punishment strategies
• Compare Boyd-Richerson model dynamics in realistic LLM simulations
• Test norm emergence via natural language reasoning in multi-agent systems

Innovation

• Using LLM agents for realistic social simulations
• Implementing explicit punishment to drive cooperation
• Integrating natural language reasoning in MAS