Simulating Opinion Dynamics with Networks of LLM-based Agents

By Yun-Shiuan Chuang, Agam Goyal, Nikunj Harlalka, Siddharth Suresh, R. Hawkins, Sijia Yang, Dhavan V. Shah, Junjie Hu, Timothy T. Rogers. 2023

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Tracking the perspectives of interacting language models (2024, background): “Recent work has included objective-less behavioral studies (Park et al., 2023), studying the formation of social networks (Papachristou and Yuan, 2024), tracking opinion dynamics via classification of LLM response (Chuang et al., 2023), and analyzing document collaboration (Perez et al., 2024).”
Control Charts for Multi-agent Systems (2026, unknown): “Current approaches for measuring system behavior are limited to aggregate agent-level statistics or qualitative inspection [7, 19].”
Detecting Perspective Shifts in Multi-agent Systems (2025, unknown): “Broadly, prior work simulates predefined agent architectures interacting within sandboxed environments (Park et al., 2023; Piao et al., 2025) or along fixed interaction graphs (Papachris-tou & Yuan, 2024; Helm et al., 2024; Chuang et al., 2024).”
When LLMs Play the Telephone Game: Cultural Attractors as Conceptual Tools to Evaluate LLMs in Multi-turn Settings (2024, unknown): “Strong, but non-human-like biases for producing factual information have also been observed (Chuang et al., 2023), stressing the importance of understanding the evolution of content in LLMs and the ways it might deviate from human cultural evolution.”