Elamur Research⌕
☰
home / papers / 2026 / SPAGBias: Uncovering and Tracing Structured Spatial Gender Bias in Large Language Models

SPAGBias: Uncovering and Tracing Structured Spatial Gender Bias in Large Language Models

Paper · 2026 · Binxian Su, Haoye Lou, Shucheng Zhu et al.

Summary

LLMs are increasingly used in urban planning and other spatial tasks. Feminist geography holds that everyday places carry gendered meanings, so these models may reproduce or amplify that hierarchy. The authors present SPAGBias, described as the first systematic approach to quantifying how LLMs associate gender with space. It has three parts:

  • Taxonomy: 62 urban micro-spaces.
  • Prompt library: a forced-choice prompt plus prompts for single-gender and mixed-gender short stories.
  • Three diagnostic layers: explicit bias from repeated sampling, probabilistic bias from token log-probabilities, and constructional bias from semantic-role and narrative-role analysis of generated stories.

Across six models, gender-space links proved systematic and finer-grained than a simple split where men belong in public and women at home. Tracing experiments covered pre-training data, instruction tuning and reward modeling. They suggest the bias is partly reduced along this pipeline but never removed, and model associations go well beyond real-world distributions. Downstream tests show two failures: stereotyped outputs in value-based tasks, and poor reflection of actual statistics in fact-based tasks. The work brings sociological theory of space into bias research.

Key points

  • The taxonomy lists 62 urban micro-spaces: 43 public (transport, leisure, commerce, healthcare, religious sites) and 19 private (kitchen, laundry room, game room, etc.).
  • Explicit bias is measured with 30 forced-choice samples per space at temperature 1, analysed with binomial tests and an EDI metric. Story analysis uses OR values, SRL and gendered-role analysis.
  • Bias tracing used the C4 corpus as a stand-in for pre-training data, along with instruction-tuning and reward-modeling stages. The authors present this as evidence of trends, not causal proof for any specific model.
  • In the descriptive downstream UP Task, accuracy was only 5% (Deepseek), 20% (Qwen) and 13.5% (GPT). Models mostly gave gender-neutral answers instead of real-world tendencies.
  • Limitations: the study covers English only, uses a binary gender framing, and includes urban spaces only (no suburban or rural ones).

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 7The 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 worksSocial Sciences · 1,772 worksSafety Research · 543 worksGender Studies · 52 works

🧭 Topics 2Research subjects and languages the work is about.Language models & pretraining · 3,320 worksBias, fairness & safety · 517 works

Cite

Cite as Binxian Su, Haoye Lou, Shucheng Zhu, Weikang Wang, Ying Liu, Dong Yu and Pengyuan Liu. 2026. SPAGBias: Uncovering and Tracing Structured Spatial Gender Bias in Large Language Models. 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.55

@inproceedings{su2026spagbias,
  title = {{SPAGBias: Uncovering and Tracing Structured Spatial Gender Bias in Large Language Models}},
  author = {Binxian Su, Haoye Lou, Shucheng Zhu, Weikang Wang, Ying Liu, Dong Yu and Pengyuan Liu},
  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.55},
  url = {https://doi.org/10.18653/v1/2026.acl-long.55}
}

References

63 references, 14 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Marah I Abdin, Sam Ade Jacobs, Ammar Ahmad Awan, Jyoti Aneja, Ahmed Awadallah, Hany Hassan Awadalla, Nguyen Bach, Amit Bahree, Arash Bakhtiari, Harkirat Behl, Alon Benhaim, Misha Bilenko, Johan Bjorck, Sébastien Bubeck, Martin Cai, Caio César Teodoro Mendes, Weizhu Chen, Vishrav Chaudhary, Parul Cho… — Phi-3 Technical Report: A Highly Capable Language Model Locally on Your Phone
  2. Alibaba Group. 2024. Qwen2: A high-performance language model family. Accessed: 2025-04-21. find
  3. Michael GW Bamberg. 1997. Positioning between structure and performance. find
  4. Yasminah Beebeejaun. 2017. Gender, urban space, and the right to everyday life. Journal of Urban Affairs, 39(3):323–334. find
  5. Emily M. Bender, Timnit Gebru, Angelina McMillan-Major, and Shmargaret Shmitchell. 2021. On the dangers of stochastic parrots: Can language models be too big? In Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency, FAccT ’21, page 610–623, New York, NY, USA. Associat… find
  6. Kirti Bhagat, Kinshuk Vasisht, and Danish Pruthi. 2025. Richer output for richer countries: Uncovering geographical disparities in generated stories and travel recommendations. Preprint, arXiv:2411.07320. find
  7. Tolga Bolukbasi, Kai-Wei Chang, James Y Zou, Venkatesh Saligrama, and Adam T Kalai. 2016. Man is to computer programmer as woman is to homemaker? debiasing word embeddings. Advances in neural information processing systems, 29.
  8. Pierre Bourdieu. 2001. Masculine domination. Stanford University Press. find
  9. Peter I Buerhaus. 2010. American nursing: A history of knowledge, authority, and the meaning of work. JAMA, 304(20):2301–2302. find
  10. Matthew Carmona. 2021. Public places urban spaces: The dimensions of urban design. Routledge. find
  11. Aili Chen, Xuyang Ge, Ziquan Fu, Yanghua Xiao, and Jiangjie Chen. 2024. Travelagent: An ai assistant for personalized travel planning. Preprint, arXiv:2409.08069. — [arXiv:2409.08069](https://arxiv.org/abs/2409.08069) find
  12. Robert William Connell. 2020. Masculinities. Routledge. find
  13. DeepSeek. 2024. Deepseek llm: Scaling open-source language models with dense and mixture-of-experts models. Accessed: 2025-04-21. find
  14. Dorothy Dinnerstein. 1999. The mermaid and the minotaur: Sexual arrangements and human malaise. Other Press, LLC. find
  15. Jesse Dodge, Maarten Sap, Ana Marasović, William Agnew, Gabriel Ilharco, Dirk Groeneveld, Margaret Mitchell, and Matt Gardner. 2021. Documenting large webtext corpora: A case study on the colossal clean crawled corpus. In Proceedings of the 2021 find
  16. Conference on Empirical Methods in Natural Language Processing, pages 1286–1305, Online and Punta Cana, Dominican Republic. Association for Computational Linguistics. find
  17. Yanai Elazar, Akshita Bhagia, Ian Magnusson, Abhilasha Ravichander, Dustin Schwenk, Alane Suhr, Pete Walsh, Dirk Groeneveld, Luca Soldaini, Sameer Singh, Hanna Hajishirzi, Noah A. Smith, and Jesse Dodge. 2024. What’s in my big data? Preprint, arXiv:2310.20707. find
  18. Jean Bethke Elshtain. 2020. Public man, private woman: Women in social and political thought. Princeton University Press. find
  19. Federico Errica, Davide Sanvito, Giuseppe Siracusano, and Roberto Bifulco. 2025. What did I do wrong? quantifying LLMs’ sensitivity and consistency to prompt engineering. In Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Hu…
  20. Betty Friedan. 2013. The feminine mystique. WW Norton & Company. find
  21. Nikhil Garg, Londa Schiebinger, Dan Jurafsky, and James Zou. 2018. Word embeddings quantify 100 years of gender and ethnic stereotypes. Proceedings of the National Academy of Sciences, 115(16). find
  22. Fabrizio Gilardi, Meysam Alizadeh, and Maël Kubli. 2023. Chatgpt outperforms crowd workers for text-annotation tasks. Proceedings of the National Academy of Sciences, 120(30):e2305016120.
  23. Erving Goffman. 1981. Forms of talk. University of Pennsylvania Press. find
  24. Aaron Grattafiori, Abhimanyu Dubey, Abhinav Jauhri, Abhinav Pandey, Abhishek Kadian, Ahmad Al-Dahle, Aiesha Letman, Akhil Mathur, Alan Schelten, Alex Vaughan, Amy Yang, Angela Fan, Anirudh Goyal, Anthony Hartshorn, Aobo Yang, Archi Mitra, Archie Sravankumar, Artem Korenev, Arthur Hinsvark, and 542 o… — The Llama 3 Herd of Models
  25. Groq. 2024. Groq api and lpu inference engine. Accessed: 2025-04-18. find
  26. Michael Alexander Kirkwood Halliday and Christian MIM Matthiessen. 2014. Halliday’s introduction to functional grammar. (No Title). find
  27. Rom Harré and 1 others. 2012. Positioning theory: Moral dimensions of social-cultural psychology. The Oxford handbook of culture and psychology, 1:191– 206. find
  28. Alibaba Cloud Intelligence. 2024. Alibaba cloud ai computing platform - qwen2 deployment. Accessed: 2025-04-18. find
  29. Susan Jamieson. 2004. Likert scales: How to (ab) use them? Medical education, 38(12):1217–1218. find
  30. Lucinda J Kaukas. 2002. Gender space architecture: An interdisciplinary introduction. find
  31. Svetlana Kiritchenko and Saif M. Mohammad. 2018. Examining gender and race bias in two hundred sentiment analysis systems. Preprint, arXiv:1805.04508.
  32. Henri Lefebrve. 1991. The production of space. Basil Backwell. find
  33. Zhonghang Li, Lianghao Xia, Xubin Ren, Jiabin Tang, Tianyi Chen, Yong Xu, and Chao Huang. 2025. Urban computing in the era of large language models. Preprint, arXiv:2504.02009. — [arXiv:2504.02009](https://arxiv.org/abs/2504.02009) find
  34. Kevin Lynch. 1964. The image of the city. MIT press. find
  35. Rohin Manvi, Samar Khanna, Marshall Burke, David Lobell, and Stefano Ermon. 2024a. Large language models are geographically biased. arXiv preprint arXiv:2402.02680. find
  36. Rohin Manvi, Samar Khanna, Gengchen Mai, Marshall Burke, David B Lobell, and Stefano Ermon. 2024b. Geollm: Extracting geospatial knowledge from large language models. In ICLR. find
  37. Doreen Massey. 2013. Space, place and gender. John Wiley & Sons. find
  38. Chandler May, Alex Wang, Shikha Bordia, Samuel R. Bowman, and Rachel Rudinger. 2019. On measuring social biases in sentence encoders. In Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and…
  39. Mintel. 2012. Women more likely to visit a salon, but a growing number of men interested in these services. Accessed: 2026-04-13. find
  40. Shujaat Mirza, Bruno Coelho, Yuyuan Cui, Christina Pöpper, and Damon McCoy. 2024. Global-liar: Factuality of llms over time and geographic regions. arXiv preprint arXiv:2401.17839. — [arXiv:2401.17839](https://arxiv.org/abs/2401.17839) find
  41. OpenAI. 2022. Gpt-3.5-turbo. Accessed: 2025-04-18. find
  42. OpenAI. 2023. Gpt-4 technical report. Preprint, arXiv:2303.08774. This report covers the development and evaluation of GPT-4.
  43. OpenAI. 2024. Openai api documentation. Accessed: 2025-04-18. find
  44. Rachel Pain. 2001. Gender, race, age and fear in the city. Urban Studies, 38(5-6):899–913. find
  45. C.C. Perez. 2019. Invisible Women: Data Bias in a World Designed for Men. Abrams Press. find
  46. Lakshmi Puri. 2016. Speech by deputy executive director lakshmi puri at the women’s and youth assembly at habitat iii. https://www.unwomen.org/en/news-stories. Speech at the Women’s and Youth Assembly, Habitat III, UN Women. ↗
  47. Matthew Renze. 2024. The effect of sampling temperature on problem solving in large language models. In Findings of the Association for Computational Linguistics: EMNLP 2024, pages 7346–7356, Miami, Florida, USA. Association for Computational Linguistics.
  48. Melanie Sclar, Yejin Choi, Yulia Tsvetkov, and Alane Suhr. 2024. Quantifying language models’ sensitivity to spurious features in prompt design or: How i learned to start worrying about prompt formatting. Preprint, arXiv:2310.11324. find
  49. Joanne Sharp. 2009. Geography and gender: what belongs to feminist geography? emotion, power and change. Progress in human geography, 33(1):74–80. find
  50. Emily Sheng, Kai-Wei Chang, Premkumar Natarajan, and Nanyun Peng. 2019. The woman worked as a babysitter: On biases in language generation. In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Process…
  51. Rachel Silvey. 2006. Geographies of gender and migration: Spatializing social difference 1. International Migration Review, 40(1):64–81. find
  52. Sharlene Mollett and Caroline Faria. 2018. The spa-tialities of intersectional thinking: fashioning feminist geographic futures. Gender, Place & Culture, 25(4):565–577. find
  53. Bernd Carsten Stahl and Damian Eke. 2024. The ethics of chatgpt – exploring the ethical issues of an emerging technology. International Journal of Information Management, 74:102700. find
  54. Morningstar. 2023. 100 must-know statistics about long-term care: 2023 edition. Accessed: 2026-04-13. find
  55. National Center for Health Statistics. 2024. NCHS publishes new data brief on residential care community resident characteristics. Accessed: 2026-04-13. find
  56. Gabriel Stanovsky, Noah A. Smith, and Luke Zettlemoyer. 2019. Evaluating gender bias in machine translation. In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, pages 1679–1684, Florence, Italy. Association for Computational Linguistics.
  57. Daisuke Oba, Masahiro Kaneko, and Danushka Bollegala. 2024. In-contextual gender bias suppression for large language models. EACL (Findings), pages 1722–1742. find
  58. Petter Törnberg. 2023. Chatgpt-4 outperforms experts and crowd workers in annotating political twitter messages with zero-shot learning. arXiv preprint arXiv:2304.06588.
  59. UN Women. 2017. Redistribute unpaid work. Accessed: 2025-04-18. find
  60. United Nations Industrial Development Organization. 2020. Women in industry: Why we need to improve statistics on gender. Accessed: 2026-04-13. find
  61. Eva Vanmassenhove, Christian Hardmeier, and Andy Way. 2018. Getting gender right in neural machine translation. In Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing, pages 3003–3008, Brussels, Belgium. Association for Computational Linguistics.
  62. Yixin Wan, George Pu, Jiao Sun, Aparna Garimella, Kai-Wei Chang, and Nanyun Peng. 2023. " kelly is a warm person, joseph is a role model": Gender biases in llm-generated reference letters. arXiv preprint arXiv:2310.09219.
  63. Angelina Wang, Michelle Phan, Daniel E. Ho, and Sanmi Koyejo. 2025. Fairness through difference awareness: Measuring Desired group discrimination in LLMs. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 6867–6893, Vienna, Aust… find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/su-2026-spagbias-uncovering-tracing-structured-spatial/