Collective Opinion Dynamics in Structured LLM Populations

arXiv:2604.11312 · cs.SI, cs.AI, cs.CY, cs.MA, physics.soc-ph · Submitted 2026-04-13 · Read on arXiv

cs.SI, cs.AI, cs.CY, cs.MA, physics.soc-ph

Submitted: 2026-04-13

Updated: 2026-09-28

Code: https://github.com/meta-llama/llama3

Project page: https://andreafailla.github.io

Terminology

• Large Language Models (LLMs) — 1× in this paper · explained in 6 episodes

Sources

• OpenAI GPT-5 System Card
• Mistral 7B
• All That's 'Human' Is Not Gold: Evaluating Human Evaluation of Generated Text
• Simulating Social Media Using Large Language Models to Evaluate Alternative News Feed Algorithms
• Characteristics and prevalence of fake social media profiles with AI-generated faces
• Emergence of Scale-Free Networks in Social Interactions among Large Language Models
• Simulating Opinion Dynamics with Networks of LLM-based Agents
• LLM Agents Grounded in Self-Reports Enable General-Purpose Simulation of Individuals
• Systematic Biases in LLM Simulations of Debates
• When Your AI Agent Succumbs to Peer-Pressure: Studying Opinion-Change Dynamics of LLMs
• SocioVerse: A World Model for Social Simulation Powered by LLM Agents and A Pool of 10 Million Real-World Users
• Gender Dynamics and Homophily in a Social Network of LLM Agents
• S$^3$: Social-network Simulation System with Large Language Model-Empowered Agents
• On the Principles behind Opinion Dynamics in Multi-Agent Systems of Large Language Models

Related papers

• Linking Scalar-Intensity Language to Structural Polarization with Validated Signed-Network Measures
• Detection and Characterization of Coordinated Online Behavior: A Survey
• Omega-N: Interpretable Structural Node Descriptors and Their Applicability Domain
• Semantic Communities and Boundary-Spanning Lyrics in K-pop: A Graph-Based Unsupervised Analysis
• Simplify to Amplify: Achieving Information-Theoretic Bounds with Fewer Steps in Spectral Community Detection
• Transmission Neural Networks: Inhibitory and Excitatory Connections