Elamur Research⌕
☰
home / papers / 2026 / Synthia: Scalable Grounded Persona Generation from Social Media Data

Synthia: Scalable Grounded Persona Generation from Social Media Data

Paper · 2026 · Vahid Rahimzadeh, Erfan Moosavi Monazzah, Mohammad Taher Pilehvar et al.

Summary

Persona-based LLM simulations in computational social science are only as trustworthy as the personas behind them. Existing approaches force a trade-off. Interview-derived personas are realistic but costly, while fully synthetic ones scale easily but contain artifacts and contradictions. SYNTHIA addresses this by building first-person life narratives from real public Bluesky posts. An LLM writes the narrative, and separate models conditioned on it answer demographic and opinion questions. The synthetic population is then reweighted to match real survey respondents. Its answers are compared statistically with human opinion distributions from the American Trends Panel, so evaluation does not depend on LLM judges. Across many configurations, SYNTHIA aligns more closely with human opinions than the prior state of the art while using models less than half the size. It also shows smaller performance gaps across demographic groups. Because each persona comes from a real account, the follower graph is preserved, which enables network-aware studies such as homophily analysis. The authors attribute most of the gains to fewer internal factual contradictions within each narrative.

Key points

  • Data: Users were kept if they posted between 100 and 1,000 times over two years, leaving roughly 170M posts from about 650K users. A random 3K-user sample was drawn to match the size of the smallest baseline. Handles, URLs, PII, replies and reposts were stripped before generation.
  • Models: Personas were generated with Gemma-3-27B and the smaller Phi-4-mini. The baseline of Moon et al. (2024), originally built with Llama-3-70B, was re-run with Gemma-3-27B for a controlled comparison.
  • Opinion alignment: Personas were evaluated on ATP Waves 34 and 99 using base (non-instruction-tuned) surveyor models across 54 configurations. SYNTHIA improved over baselines by up to 11.6%.
  • Fairness: The accuracy gap between the best- and worst-performing demographic subgroups shrank by up to 25%, with lower disparate impact for gender and education.
  • Network resource: The release includes the personas and a directed follower graph under a research-only license. In the homophily case studies, link prediction improved by 8.3% (p < 0.001), and separability between connected and unconnected personas in embedding space rose by up to 46%.

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

 Institutions

🎓Cardiff UniversityCardiff · United Kingdom22 works · 2,444 citations🎓University of TehranTehran · Iran8 works · 182 citations🎓Khatam UniversityTehran · Iran4 works · 31 citations

 Fields, topics and concepts

📚 Shelves 5The field of science this work belongs to (OpenAlex fields).Computer Science · 16,667 worksArtificial Intelligence · 16,194 worksArts and Humanities · 16,039 worksLanguage and Linguistics · 15,891 worksHuman-Computer Interaction · 141 works

Cite

Cite as Vahid Rahimzadeh, Erfan Moosavi Monazzah, Mohammad Taher Pilehvar and Yadollah Yaghoobzadeh. 2026. Synthia: Scalable Grounded Persona Generation from Social Media Data. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers).

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2026.acl-long.2134

@inproceedings{rahimzadeh2026synthia,
  title = {{Synthia: Scalable Grounded Persona Generation from Social Media Data}},
  author = {Vahid Rahimzadeh, Erfan Moosavi Monazzah, Mohammad Taher Pilehvar and Yadollah Yaghoobzadeh},
  booktitle = {Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)},
  year = {2026},
  doi = {10.18653/v1/2026.acl-long.2134},
  url = {https://doi.org/10.18653/v1/2026.acl-long.2134}
}

References

51 references, 11 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Marwa Abdulhai, Ryan Cheng, Donovan Clay, Tim Althoff, Sergey Levine, and Natasha Jaques. 2025. Consistently simulating human personas with multi-turn reinforcement learning. Preprint, arXiv:2511.00222. — [arXiv:2511.00222](https://arxiv.org/abs/2511.00222) find
  2. Jacy Reese Anthis, Ryan Liu, Sean M. Richardson, Austin C. Kozlowski, Bernard Koch, James Evans, Erik Brynjolfsson, and Michael Bernstein. 2025. Llm social simulations are a promising research method. Preprint, arXiv:2504.02234. — [arXiv:2504.02234](https://arxiv.org/abs/2504.02234) find
  3. Lisa P. Argyle, Ethan C. Busby, Nancy Fulda, Joshua R. Gubler, Christopher Rytting, and David Wingate. 2023. Out of one, many: Using language models to simulate human samples. Political Analysis, 31(3):337–351.
  4. Julia Barnett, Kimon Kieslich, and Nicholas Diakopoulos. 2024. Simulating policy impacts: Developing a generative scenario writing method to evaluate the perceived effects of regulation. In Proceedings of the AAAI/ACM Conference on AI, Ethics, and Society, volume 7, pages 82–93. find
  5. Solon Barocas, Moritz Hardt, and Arvind Narayanan. 2023. Fairness and Machine Learning: Limitations and Opportunities. MIT Press. find
  6. Ngoc Bui, Hieu Trung Nguyen, Shantanu Kumar, Julian Theodore, Weikang Qiu, Viet Anh Nguyen, and Rex Ying. 2025. Mixture-of-personas language models for population simulation. In Findings of the Association for Computational Linguistics: ACL 2025, pages 24761–24778, Vienna, Austria. Association for C… find
  7. Aurélien Bück-Kaeffer, Je Qin Chooi, Dan Zhao, Maximilian Puelma Touzel, Kellin Pelrine, Jean-François Godbout, Reihaneh Rabbany, and Zachary Yang. 2025. BluePrint: A social media user dataset for llm persona evaluation and training. Preprint, arXiv:2510.02343. — [arXiv:2510.02343](https://arxiv.org/abs/2510.02343) find
  8. Hongzhan Chen, Hehong Chen, Ming Yan, Wenshen Xu, Gao Xing, Weizhou Shen, Xiaojun Quan, Chenliang Li, Ji Zhang, and Fei Huang. 2024a. Social-Bench: Sociality evaluation of role-playing conversational agents. In Findings of the Association for Computational Linguistics: ACL 2024, pages 2108–2126, Ban… find
  9. Jiangjie Chen, Xintao Wang, Rui Xu, Siyu Yuan, Yikai Zhang, Wei Shi, Jian Xie, Shuang Li, Ruihan Yang, Tinghui Zhu, Aili Chen, Nianqi Li, Lida Chen, Caiyu Hu, Siye Wu, Scott Ren, Ziquan Fu, and Yanghua Xiao. 2024b. From persona to personalization: A survey on role-playing language agents. Preprint,…
  10. Myra Cheng, Tiziano Piccardi, and Diyi Yang. 2023. CoMPosT: Characterizing and evaluating caricature in LLM simulations. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 10853–10875, Singapore. Association for Computational Linguistics. find
  11. Hyeong Kyu Choi and Yixuan Li. 2024. Picle: Eliciting diverse behaviors from large language models with persona in-context learning. In International Conference on Machine Learning, pages 8722–8739. PMLR. find
  12. Celia Cintas, Miriam Rateike, Erik Miehling, Elizabeth Daly, and Skyler Speakman. 2025. Localizing persona representations in llms. Preprint, arXiv:2505.24539. — [arXiv:2505.24539](https://arxiv.org/abs/2505.24539) find
  13. Tejaswani Dash, Dinesh Karri, Anudeep Vurity, Gautam Datla, Tazeem Ahmad, Saima Rafi, and Rohith Tangudu. 2025. Polypersona: Persona-grounded llm for synthetic survey responses. Preprint, arXiv:2512.14562. — [arXiv:2512.14562](https://arxiv.org/abs/2512.14562) find
  14. Xiangyang He, Jiale Li, Jiahao Chen, Yang Yang, and Mingming Fan. 2025. Simupanel: A novel immersive multi-agent system to simulate interactive expert panel discussion. Preprint, arXiv:2506.16010. — [arXiv:2506.16010](https://arxiv.org/abs/2506.16010) find
  15. Zihao He, Minh Duc Chu, Rebecca Dorn, Siyi Guo, and Kristina Lerman. 2024. Community-cross-instruct: Unsupervised instruction generation for aligning large language models to online communities. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, pages 17001– 1… find
  16. Or Honovich, Leshem Choshen, Roee Aharoni, Ella Neeman, Idan Szpektor, and Omri Abend. 2021. q 2 : Evaluating factual consistency in knowledge-grounded dialogues via question generation and question answering. In Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing,…
  17. EunJeong Hwang, Bodhisattwa Majumder, and Niket Tandon. 2023. Aligning language models to user opinions. In Findings of the Association for Computational Linguistics: EMNLP 2023, pages 5906– 5919, Singapore. Association for Computational Linguistics.
  18. Soon-Gyo Jung, Joni Salminen, Kholoud Khalil Aldous, and Bernard J. Jansen. 2025. Personacraft: Leveraging language models for data-driven persona development. International Journal of Human-Computer Studies, 197:103445. find
  19. Philippe Laban, Tobias Schnabel, Paul N. Bennett, and Marti A. Hearst. 2022. SummaC: Re-visiting NLI-based models for inconsistency detection in summarization. Transactions of the Association for Computational Linguistics, 10:163–177. find
  20. Ang Li, Haozhe Chen, Hongseok Namkoong, and Tianyi Peng. 2025. Llm generated persona is a promise with a catch. Preprint, arXiv:2503.16527. — [arXiv:2503.16527](https://arxiv.org/abs/2503.16527) find
  21. Andy Liu, Mona Diab, and Daniel Fried. 2024. Evaluating large language model biases in persona-steered generation. In Findings of the Association for Computational Linguistics: ACL 2024, pages 9832–9850, Bangkok, Thailand. Association for Computational Linguistics. find
  22. Genglin Liu, Vivian T. Le, Salman Rahman, Elisa Kreiss, Marzyeh Ghassemi, and Saadia Gabriel. 2025. MOSAIC: Modeling social AI for content dissemination and regulation in multi-agent simulations. In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, pages 6401–64… find
  23. Ninareh Mehrabi, Fred Morstatter, Nripsuta Saxena, Kristina Lerman, and Aram Galstyan. 2021. A survey on bias and fairness in machine learning. ACM computing surveys (CSUR), 54(6):1–35.
  24. Sewon Min, Kalpesh Krishna, Xinxi Lyu, Mike Lewis, Wen-tau Yih, Pang Koh, Mohit Iyyer, Luke Zettlemoyer, and Hannaneh Hajishirzi. 2023. FActScore: Fine-grained atomic evaluation of factual precision in long form text generation. In Proceedings of the 2023 Conference on Empirical Methods in Natural L…
  25. Suhong Moon, Marwa Abdulhai, Minwoo Kang, Joseph Suh, Widyadewi Soedarmadji, Eran Kohen Behar, and David Chan. 2024. Virtual personas for language models via an anthology of backstories. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, pages 19864–19897, Mia… find
  26. 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
  27. Joon Sung Park, Carolyn Q Zou, Aaron Shaw, Benjamin Mako Hill, Carrie Cai, Meredith Ringel Morris, Robb Willer, Percy Liang, and Michael S Bernstein. 2024. Generative agent simulations of 1,000 people. arXiv preprint arXiv:2411.10109. find
  28. Pew Research Center. 2018. American trends panel wave 34. Dataset. Field dates: April 23 – May 6, 2018. Topics: Biomedical and food issues. find
  29. Shibani Santurkar, Esin Durmus, Faisal Ladhak, Cinoo Lee, Percy Liang, and Tatsunori Hashimoto. 2023. Whose opinions do language models reflect? In Proceedings of the 40th International Conference on Machine Learning, ICML’23. JMLR.org.
  30. Jocelyn J Shen, Akhila Yerukola, Xuhui Zhou, Cynthia Breazeal, Maarten Sap, and Hae Won Park. 2025. Words like knives: Backstory-personalized modeling and detection of violent communication. In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, pages 11607–11625,… find
  31. Christine V Stephens and Mary Breheny. 2013. Narrative analysis in psychological research: An integrated approach to interpreting stories. Qualitative Research in Psychology, 10:14 – 27. find
  32. Joseph Suh, Erfan Jahanparast, Suhong Moon, Minwoo Kang, and Serina Chang. 2025a. Language model fine-tuning on scaled survey data for predicting distributions of public opinions. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), page…
  33. Pew Research Center. 2021. American trends panel wave 99. Dataset. Field dates: Nov. 1 – Nov. 7, 2021. Topics: Artificial Intelligence (AI) and human enhancement. find
  34. Joseph Suh, Suhong Moon, and Serina Chang. 2025b. Rethinking llm human simulation: When a graph is what you need. Preprint, arXiv:2511.02135. — [arXiv:2511.02135](https://arxiv.org/abs/2511.02135) find
  35. Jinghua Piao, Yuwei Yan, Jun Zhang, Nian Li, Junbo Yan, Xiaochong Lan, Zhihong Lu, Zhiheng Zheng, Jing Yi Wang, Di Zhou, Chen Gao, Fengli Xu, Fang Zhang, Ke Rong, Jun Su, and Yong Li. 2025. Agentsociety: Large-scale simulation of llm-driven generative agents advances understanding of human behaviors… — [arXiv:2502.08691](https://arxiv.org/abs/2502.08691) find
  36. Lipeipei Sun, Tianzi Qin, Anran Hu, Jiale Zhang, Shuo-jia Lin, Jianyan Chen, Mona Ali, and Mirjana Prpa. 2025. Persona-l has entered the chat: Leveraging llms and ability-based framework for personas of people with complex needs. In Proceedings of the 2025 CHI Conference on Human Factors in Computin… find
  37. Giorgio Piatti, Zhijing Jin, Max Kleiman-Weiner, Bernhard Schölkopf, Mrinmaya Sachan, and Rada Mihalcea. 2024. Cooperate or collapse: Emergence of sustainable cooperation in a society of llm agents. Advances in Neural Information Processing Systems, 37:111715–111759. find
  38. Nusrat Jahan Prottasha, Md Kowsher, Hafijur Raman, Israt Jahan Anny, Prakash Bhat, Ivan Garibay, and Ozlem Garibay. 2025. User profile with large language models: Construction, updating, and benchmarking. Preprint, arXiv:2502.10660. — [arXiv:2502.10660](https://arxiv.org/abs/2502.10660) find
  39. Vahid Rahimzadeh, Ali Hamzehpour, Azadeh Shakery, and Masoud Asadpour. 2025. From millions of tweets to actionable insights: Leveraging llms for user profiling. arXiv preprint arXiv:2505.06184. — [arXiv:2505.06184](https://arxiv.org/abs/2505.06184) find
  40. Leonard Salewski, Stephan Alaniz, Isabel Rio-Torto, Eric Schulz, and Zeynep Akata. 2023. In-context impersonation reveals large language models’ strengths and biases. In Proceedings of the 37th International Conference on Neural Information Processing Systems, NIPS ’23, Red Hook, NY, USA. Curran Ass… find
  41. Gemma Team, Aishwarya Kamath, Johan Ferret, Shreya Pathak, Nino Vieillard, Ramona Merhej, Sarah Perrin, Tatiana Matejovicova, Alexandre Ramé, Morgane Rivière, Louis Rouillard, Thomas Mesnard, Geoffrey Cideron, Jean bastien Grill, Sabela Ramos, Edouard Yvinec, Michelle Casbon, Etienne Pot, Ivo Penche… find
  42. Microsoft Team, Abdelrahman Abouelenin, Atabak Ashfaq, Adam Atkinson, Hany Awadalla, Nguyen Bach, Jianmin Bao, Alon Benhaim, Martin Cai, Vishrav Chaudhary, Congcong Chen, Dong Chen, Dongdong Chen, Junkun Chen, Weizhu Chen, Yen-Chun Chen, Yi ling Chen, Qi Dai, Xiyang Dai, and 56 others. 2025b. Phi-4-…
  43. Maximilian Puelma Touzel, Sneheel Sarangi, Austin Welch, Gayatri Krishnakumar, Dan Zhao, Zachary Yang, Hao Yu, Ethan Kosak-Hine, Tom Gibbs, Andreea Musulan, Camille Thibault, Busra Tugce Gurbuz, Reihaneh Rabbany, Jean-François Godbout, • Phi-4-mini-instruct 4B: Utilized for both persona generation f… — [arXiv:2410.13915](https://arxiv.org/abs/2410.13915) find
  44. Yu-Min Tseng, Yu-Chao Huang, Teng-Yun Hsiao, Wei-Lin Chen, Chao-Wei Huang, Yu Meng, and Yun-Nung Chen. 2024. Two tales of persona in LLMs: A survey of role-playing and personalization. In Findings of the Association for Computational Linguistics: EMNLP 2024, pages 16612–16631, Miami, Florida, USA. A…
  45. Gemini 2.5 Flash & Claude 3.7 Sonnet & GPT-4.1: Used for the inconsistency detection pipeline, accessed through the available APIs on Openrouter.ai platform. • ModernBERT-base & Qwen3-Embedding-0.6B: Used for the link prediction and distribution analysis respectively. find
  46. Kuang Wang, Xianfei Li, Shenghao Yang, Li Zhou, Feng Jiang, and Haizhou Li. 2025a. Know you first and be you better: Modeling human-like user simulators via implicit profiles. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 21… find
  47. Lei Wang, Jingsen Zhang, Hao Yang, Zhi-Yuan Chen, Jiakai Tang, Zeyu Zhang, Xu Chen, Yankai Lin, Hao Sun, Ruihua Song, Xin Zhao, Jun Xu, Zhicheng Dou, Jun Wang, and Ji-Rong Wen. 2025b. User behavior simulation with large language model-based agents. ACM Trans. Inf. Syst., 43(2). find
  48. Dulce Wilkinson Westberg, Moin Syed, Aerika Brittian Loyd, and William Dunlop. 2024. Using intersectionality to understand how structural domains are embedded in life narratives. Journal of personality. find
  49. Rui Xu, Xintao Wang, Jiangjie Chen, Siyu Yuan, Xinfeng Yuan, Jiaqing Liang, Zulong Chen, Xiaoqing Dong, and Yanghua Xiao. 2024. Character is destiny: Can role-playing language agents make persona-driven decisions? Preprint, arXiv:2404.12138. — [arXiv:2404.12138](https://arxiv.org/abs/2404.12138) find
  50. Min Yin, Haoyu Liu, Boyi Lian, and Chunlei Chai. 2025. Co-persona: Leveraging llms and expert collaboration to understand user personas through social media data analysis. Preprint, arXiv:2506.18269. — [arXiv:2506.18269](https://arxiv.org/abs/2506.18269) find
  51. Saizheng Zhang, Emily Dinan, Jack Urbanek, Arthur Szlam, Douwe Kiela, and Jason Weston. 2018. Personalizing dialogue agents: I have a dog, do you have pets too? In Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 2204–2213, Melbou…

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/rahimzadeh-2026-synthia-scalable-grounded-persona-generation/