Skip to content

Agent-based Modeling: Equilibrium, Echo Chambers, and Efficiency in Hybrid Coevolutionary Opinion Games

Sep 2026 · arXiv.org · Vol abs/2609.27639 · 1 citation · 62 references
Computer Science

TL;DR

The Hybrid Coevolutionary Opinion Game (H-COG) is introduced, in which analytical and LLM-driven agents share one network, choose their neighbors by opinion similarity in every round, and hold stances drawn from real Reddit comments on gun control and abortion.

Abstract

Online discussion of political and gender-related issues is often heated, and when opinions in a network draw closer, the convergence is readily taken as genuine consensus. Whether it carries a cost is a question existing methods cannot answer: coevolutionary opinion formation games measure the Price of Anarchy (PoA) of agents that update by numerical rules, while simulations with large language model (LLM) agents report only descriptive indices. We introduce the Hybrid Coevolutionary Opinion Game (H-COG), in which analytical and LLM-driven agents share one network, choose their neighbors by opinion similarity in every round, and hold stances drawn from real Reddit comments on gun control and abortion. To our knowledge, H-COG is the first framework to place Friedkin-Johnsen best-response agents and LLM agents in one coevolutionary game and to measure the social cost and PoA of LLM-driven populations. We prove that on any fixed network, given the LLM agents'opinions, the analytical agents'opinion stage has a unique equilibrium and the social optimum has a closed form, and that the convergence guarantee of Chen et al. for optimistic gradient ascent carries over to H-COG. All runs converge structurally. LLM-driven populations are less polarized yet have about five times the PoA of analytical ones; half of the gap comes from agents being pulled away from their own prior positions, a distance we prove must carry a cost whenever expressed opinions are more concentrated than intrinsic ones. Echo chambers form under every composition and grow out of the rewiring rule rather than the initial topology. Opinions in these populations draw closer largely because agents give up their own positions.

View source

Similar papers

Open access Sep 2026

Research on the Game Dynamics of Optional Public Goods

The main findings demonstrate that individuals update strategies primarily through self-adjustment based on historical payoffs, with imitation playing merely an auxiliary role and the optimal self-adjustment proportion is approximately 0.9, and low sensitivity coefficients and mutation rates favor the emergence of coop...

Hao-Chen Wu, Meng-Cheng Sun, Lu-He Yang et al. · 0 citations
#machine learning Preprint Sep 2026

Be Careful Who You Trust: Coordination Dynamics under Corrupted Communication in LLM Multi-Agent Games

The results show that original choices, public actions, and executed outcomes must be separated when evaluating multi-agent robustness, as corrupted communication can severely and predictably degrade mutually beneficial cooperation.

Xuan-Yi Liu, Niall J. Dalton, Hairil Amin et al. · 0 citations
Preprint Sep 2026

Cheap Talk Stabilizes Strategic Interaction in LLM Agents

Large language models are increasingly deployed as interacting agents, making the persistence of their action policies across repeated interaction critical for reliable multi-agent operation. We investigate whether and how agent-generated, non-binding pre-play communication ("cheap talk") increases such persistence in...

Nunzio Lorè, Hon-Gan Zhu, Babak Heydari · 0 citations
Preprint Oct 2026

Cooperation, Defection, and the Commons: Reproducing an Agent-Based Model of Aksum with Three Logistic Equations

Cooperative behavior can improve environmental quality, whereas selfish defection can degrade shared resources. In a recent agent-based study of the Aksum civilization (Vasellini et al., 2025), a population of agents moving over a reconstructed landscape and playing a local Public Goods Game grew faster and peaked high...

Riccardo Vasellini, F. Cecconi, Daniele Vilone · 0 citations

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.