Elamur Research⌕
☰
home / papers / 2025 / Decoding Echo Chambers: LLM-Powered Simulations Revealing Polarization in Social Networks

Decoding Echo Chambers: LLM-Powered Simulations Revealing Polarization in Social Networks

Paper · 2025 · Chenxi Wang, Zongfang Liu, Dequan Yang et al.

Summary

Classic opinion-dynamics models such as the Friedkin-Johnsen (FJ) model and the Bounded Confidence Model (BCM) represent people's views as plain numbers. They therefore miss the fact that persuasion on social media happens through text. The authors introduce a Social Simulation Framework (SSF) in which every user is an LLM agent with:

  • a persona (age, gender, education, traits),
  • short- and long-term memory,
  • a reflection step.

Agents are placed on small-world, scale-free, or random graphs. They talk only with network neighbours, and a recommender-style rule favours neighbours with similar views. Each simulated day, an agent writes an opinion and assigns itself a belief score from -2 to 2. The authors compare SSF with BCM and FJ using three measures: neighbor correlation index (NCI), polarization, and global disagreement. Echo chambers appear on small-world and scale-free graphs but not on random ones, and SSF reproduces polarization more convincingly than the numeric baselines. The authors also test two text-based interventions, active and passive nudges, which lower both polarization and NCI. The work offers an explainable, language-grounded tool for studying online polarization.

Key points

  • Three graph types are tested: Watts–Strogatz small-world, Barabási–Albert scale-free, and random graphs. Echo chambers form only on the first two, while random networks show rising global disagreement instead.
  • Agents' Big Five traits behave plausibly. "Credulous" agents (high Agreeableness/Neuroticism) shifted their beliefs more on average (mean 1.0909, variance 0.8056) than "Skeptical" ones (0.8333, 0.8099).
  • Active and passive nudges, applied on a scale-free network, reduced extreme opinions, NCI, and the polarization metric. A case study of one agent's views on euthanasia shows slower, milder escalation when nudges are applied.
  • Limitations include a network size far below real platforms and topic biases inherited from the LLM.
  • Code is released as EchoChamberSim on GitHub (ZongfangLiu/EchoChamberSim).

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Fields, topics and concepts

📚 Shelves 5The field of science this work belongs to (OpenAlex fields).Social Sciences · 1,772 worksSociology and Political Science · 577 worksPhysics and Astronomy · 44 worksStatistical and Nonlinear Physics · 40 worksCommunication · 97 works

🧭 Topics 1Research subjects and languages the work is about.Language models & pretraining · 3,320 works

Cite

Cite as Chenxi Wang, Zongfang Liu, Dequan Yang and Xiuying Chen. 2025. Decoding Echo Chambers: LLM-Powered Simulations Revealing Polarization in Social Networks. Proceedings of the 31st International Conference on Computational Linguistics.

The paper itself, not this page. BibTeX from the library's record of the work.

@inproceedings{wang2025decoding,
  title = {{Decoding Echo Chambers: LLM-Powered Simulations Revealing Polarization in Social Networks}},
  author = {Chenxi Wang, Zongfang Liu, Dequan Yang and Xiuying Chen},
  booktitle = {Proceedings of the 31st International Conference on Computational Linguistics},
  year = {2025},
  url = {https://aclanthology.org/2025.coling-main.264/}
}

References

38 references, 1 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Faisal Alatawi, Lu Cheng, Anique Tahir, Mansooreh Karami, Bohan Jiang, Tyler Black, and Huan Liu. 2021. A survey on echo chambers on social media: Description, detection and mitigation. arXiv preprint arXiv:2112.05084. — [arXiv:2112.05084](https://arxiv.org/abs/2112.05084) find
  2. Albert-László Barabási and Réka Albert. 1999. Emergence of scaling in random networks. science, pages 509–512. find
  3. Murray R Barrick and Michael K Mount. 1991. The big five personality dimensions and job performance: a meta-analysis. Personnel psychology, pages 1–26. find
  4. Alessandro Bessi, Fabio Petroni, Michela Del Vicario, Fabiana Zollo, Aris Anagnostopoulos, Antonio Scala, Guido Caldarelli, and Walter Quattrociocchi. 2016. Homophily and polarization in the age of misinformation. The European physical journal special topics, pages 2047–2059. find
  5. Uthsav Chitra and Christopher Musco. 2020a. Analyzing the impact of filter bubbles on social network polarization. In Proceedings of the 13th International Conference on Web Search and Data Mining, pages 115–123. find
  6. Uthsav Chitra and Christopher Musco. 2020b. Analyzing the impact of filter bubbles on social network polarization. In Proceedings of the 13th International Conference on Web Search and Data Mining, page 115–123. find
  7. Yun-Shiuan Chuang, Agam Goyal, Nikunj Harlalka, Siddharth Suresh, Robert Hawkins, Sijia Yang, Dha-In our work, we consider diverse graph structures van Shah, Junjie Hu, and Timothy Rogers. 2024. Simto mimic the real-world opinion propagation pro-ulating opinion dynamics with networks of llm-based ce… find
  8. Federico Cinus, Marco Minici, Corrado Monti, and Francesco Bonchi. 2021. The effect of people recommenders on echo chambers and polarization. In International Conference on Web and Social Media. find
  9. Yize Li, Jiazhong Nie, Yi Zhang, Bingqing Wang, Baoshi Yan, and Fuliang Weng. 2010. Contextual recommendation based on text mining. In Proc. of COLING, pages 692–700. find
  10. Peter M Dahlgren. 2020. Media echo chambers: Selective exposure and confirmation bias in media use, and its consequences for political polarization. find
  11. Yuhan Liu, Xiuying Chen, Xiaoqing Zhang, Xing Gao, Ji Zhang, and Rui Yan. 2024a. From skepticism to acceptance: Simulating the attitude dynamics toward fake news. IJCAI. find
  12. Guillaume Deffuant, David Neau, Frederic Amblard, and Gérard Weisbuch. 2000. Mixing beliefs among interacting agents. Advances in Complex Systems, pages 87–98. find
  13. P ERDdS and A R&wi. 1959. On random graphs i. Publ. math. debrecen, page 18. find
  14. Noah E Friedkin and Eugene C Johnsen. 1990. Social influence and opinions. Journal of mathematical sociology, pages 193–206. find
  15. King-wa Fu, Chung-hong Chan, and Michael Chau. 2013. Assessing censorship on microblogs in china: Discriminatory keyword analysis and the real-name registration policy. IEEE internet computing, pages 42–50. find
  16. Nabeel Gillani, Ann Yuan, Martin Saveski, Soroush Vosoughi, and Deb Roy. 2018. Me, my echo chamber, and i: introspection on social media polarization. In Proc. of WWW, pages 823–831. find
  17. Max Grömping. 2014. ‘echo chambers’ partisan facebook groups during the 2014 thai election. Asia Pacific Media Educator, pages 39–59. find
  18. Taicheng Guo, Xiuying Chen, Yaqi Wang, Ruidi Chang, Shichao Pei, Nitesh V Chawla, Olaf Wiest, and Xiangliang Zhang. 2024. Large language model based multi-agents: A survey of progress and challenges. IJCAI.
  19. Nada Ibrahim, Mariam Elzayany, and Amr Elmougy. 2022. The effects of personality traits on rumors. find
  20. Zhao Kaiya, Michelangelo Naim, Jovana Kondic, Manuel Cortes, Jiaxin Ge, Shuying Luo, Guangyu Robert Yang, and Andrew Ahn. 2023. Lyfe agents: Generative agents for low-cost real-time social interactions. arXiv preprint arXiv:2310.02172. — [arXiv:2310.02172](https://arxiv.org/abs/2310.02172) find
  21. Jackie Kazil, David Masad, and Andrew Crooks. 2020. Utilizing python for agent-based modeling: The mesa framework. In Social, Cultural, and Behavioral Modeling: 13th International Conference, SBP-BRiMS 2020, Washington, DC, USA, October 18–21, 2020, Proceedings 13, pages 308–317. find
  22. Gilat Levy and Ronny Razin. 2019. Echo chambers and their effects on economic and political outcomes. Annual Review of Economics, pages 303–328. find
  23. Chao Li, Xing Su, Chao Fan, Haoying Han, Cong Xue, and Chunmo Zheng. 2023. Quantifying the impact of large language models on collective opinion dynamics. arXiv preprint arXiv:2308.03313. — [arXiv:2308.03313](https://arxiv.org/abs/2308.03313) find
  24. Yuhan Liu, Zirui Song, Xiaoqing Zhang, Xiuying Chen, and Rui Yan. 2024b. From a tiny slip to a giant leap: An llm-based simulation for fake news evolution. arXiv preprint arXiv:2410.19064. — [arXiv:2410.19064](https://arxiv.org/abs/2410.19064) find
  25. Mahdieh Mirzabeigi, Mahsa Torabi, and Tahere Jowkar. 2023. The role of personality traits and the ability to detect fake news in predicting information avoidance during the covid-19 pandemic. Library Hi Tech. find
  26. Matan Orbach, Yonatan Bilu, Assaf Toledo, Dan Lahav, Michal Jacovi, Ranit Aharonov, and Noam Slonim. 2020. Out of the echo chamber: Detecting countering debate speeches. In Proc. of ACL, pages 7073–7086. find
  27. Eli Pariser. 2011. The filter bubble: What the Internet is hiding from you. penguin UK. find
  28. Joon Sung Park, Joseph O’Brien, Carrie Jun Cai, Meredith Ringel Morris, Percy Liang, and Michael S Bernstein. 2023. Generative agents: Interactive simulacra of human behavior. In Proceedings of the 36th annual acm symposium on user interface software and technology, pages 1–22. find
  29. Joon Sung Park, Lindsay Popowski, Carrie Cai, Meredith Ringel Morris, Percy Liang, and Michael S Bernstein. 2022. Social simulacra: Creating populated prototypes for social computing systems. In Proceedings of the 35th Annual ACM Symposium on User Interface Software and Technology, pages 1–18. find
  30. Ludovic Terren Ludovic Terren and Rosa Borge-Bravo Rosa Borge-Bravo. 2021. Echo chambers on social media: A systematic review of the literature. Review of Communication Research. find
  31. Johan Ugander, Brian Karrer, Lars Backstrom, and Cameron Marlow. 2011. The anatomy of the facebook social graph. arXiv preprint arXiv:1111.4503. — [arXiv:1111.4503](https://arxiv.org/abs/1111.4503) find
  32. Lei Wang, Jingsen Zhang, Hao Yang, Zhiyuan Chen, Jiakai Tang, Zeyu Zhang, Xu Chen, Yankai Lin, Ruihua Song, Wayne Xin Zhao, et al. 2023. User behavior simulation with large language model based agents. arXiv preprint arXiv:2306.02552. find
  33. Xiao Fan Wang and Guanrong Chen. 2003. Complex networks: small-world, scale-free and beyond. IEEE circuits and systems magazine, pages 6–20. find
  34. Duncan J Watts and Steven H Strogatz. 1998. Collective dynamics of ‘small-world’networks. nature, pages 440–442. find
  35. Ross Williams, Niyousha Hosseinichimeh, Aritra Majumdar, and Navid Ghaffarzadegan. 2023. Epidemic modeling with generative agents. Technical report. find
  36. Chengxing Xie, Canyu Chen, Feiran Jia, Ziyu Ye, Kai Shu, Adel Bibi, Ziniu Hu, Philip Torr, Bernard Ghanem, and Guohao Li. Can large language model agents simulate human trust behaviors? In ICLR 2024 Workshop: How Far Are We From AGI. find
  37. Xiaoqing Zhang, Xiuying Chen, Yuhan Liu, Jianzhou Wang, Zhenxing Hu, and Rui Yan. 2024. Llm-driven agents for influencer selection in digital advertising campaigns. arXiv preprint arXiv:2403.15105. — [arXiv:2403.15105](https://arxiv.org/abs/2403.15105) find
  38. Xinyi Zhou and Reza Zafarani. 2020. A survey of fake news: Fundamental theories, detection methods, and opportunities. ACM Comput. Surv. find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/wang-2025-decoding-echo-chambers-llm-powered/