Marked Personas: Using Natural Language Prompts to Measure Stereotypes in Language Models
Open accessThe DOI matches its Crossref record; 30 of the 30 references it cites were found in a registry or an index.
Abstract
To recognize and mitigate harms from large language models (LLMs), we need to understand the prevalence and nuances of stereotypes in LLM outputs.Toward this end, we present Marked Personas, a prompt-based method to measure stereotypes in LLMs for intersectional demographic groups without any lexicon or data labeling.Grounded in the sociolinguistic concept of markedness (which characterizes explicitly linguistically marked categories versus unmarked defaults), our proposed method is twofold: 1) prompting an LLM to generate personas, i.e., natural language descriptions, of the target demographic group alongside personas of unmarked, default groups; 2) identifying the words that significantly distinguish personas of the target group from corresponding unmarked ones.We find that the portrayals generated by GPT-3.5 and GPT-4 contain higher rates of racial stereotypes than human-written portrayals using the same prompts.The words distinguishing personas of marked (non-white, non-male) groups reflect patterns of othering and exoticizing these demographics.An intersectional lens further reveals tropes that dominate portrayals of marginalized groups, such as tropicalism and the hypersexualization of minoritized women.These representational harms have concerning implications for downstream applications like story generation.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
VADER: a parsimonious rule-based model for sentiment analysis of social media text
Hutto, C., & Gilbert, E. (2014) · Proceedings of the International AAAI Conference on Web and Social Media · doi:10.1609/icwsm.v8i1.14550 · Cited by 6,087
Feminist Theory: From Margin to Center
Jackson, S. (1985) · Ufahamu A Journal of African Studies · doi:10.5070/f7143017029 · Cited by 4,441
Training language models to follow instructions with human feedback
Ouyang, L., Wu, J., Jiang, X., Almeida, D., Wainwright, C. L., Mishkin, P., Zhang, C., Agarwal, S., Slama, K., Ray, A., et al. (2022) · arXiv · doi:10.48550/arxiv.2203.02155 · Cited by 4,208
Semantics derived automatically from language corpora contain human-like biases
Caliskan, A., Bryson, J. J., & Narayanan, A. (2017) · Science · doi:10.1126/science.aal4230 · Cited by 2,887
Heilman, M. E. (2001) · Journal of Social Issues · doi:10.1111/0022-4537.00234 · Cited by 2,482
The Inmates are Running the Asylum
Cooper, A. (1999) · Berichte des German Chapter of the ACM · doi:10.1007/978-3-322-99786-9_1 · Cited by 1,855
Eagly, A. H., Nater, C., Miller, D. I., Kaufmann, M., & Sczesny, S. (2020) · American Psychologist · doi:10.1037/amp0000494 · Cited by 1,136
Superwoman Schema: African American Women’s Views on Stress, Strength, and Health
Woods-Giscombé, C. L. (2010) · Qualitative Health Research · doi:10.1177/1049732310361892 · Cited by 802
Liboiron, M. (2021) · doi:10.1515/9781478021445 · Cited by 716
StereoSet: Measuring stereotypical bias in pretrained language models
Nadeem, M., Bethke, A., & Reddy, S. (2021) · Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers) · doi:10.18653/v1/2021.acl-long.416 · Cited by 643
“What Group?” Studying Whites and Whiteness in the Era of “Color-Blindness”
Lewis, A. E. (2004) · Sociological Theory · doi:10.1111/j.0735-2751.2004.00237.x · Cited by 643
StereoSet: Measuring stereotypical bias in pretrained language models
Nadeem, M., Bethke, A., & Reddy, S. (2020) · arXiv · doi:10.48550/arxiv.2004.09456 · Cited by 643
Monroe, B. L., Colaresi, M. P., & Quinn, K. M. (2008) · Political Analysis · doi:10.1093/pan/mpn018 · Cited by 542
Holistic Evaluation of Language Models
Bommasani, R., Liang, P., & Lee, T. (2023) · Annals of the New York Academy of Sciences · doi:10.1111/nyas.15007 · Cited by 480
Ethical and social risks of harm from Language Models
Weidinger, L., Mellor, J., Rauh, M., Griffin, C., Uesato, J., Huang, P.-S., Cheng, M., Glaese, M., Balle, B., Kasirzadeh, A., et al. (2021) · arXiv · doi:10.48550/arxiv.2112.04359 · Cited by 477
Self-citationOn Measuring Social Biases in Sentence Encoders
May, C., Wang, A., Bordia, S., Bowman, S. R., & Rudinger, R. (2019) · Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers) · doi:10.18653/v1/n19-1063 · Cited by 461
CrowS-Pairs: A Challenge Dataset for Measuring Social Biases in Masked Language Models
Nangia, N., Vania, C., Bhalerao, R., & Bowman, S. R. (2020) · Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.18653/v1/2020.emnlp-main.154 · Cited by 446
Lee, M., Liang, P., & Yang, Q. (2022) · CHI Conference on Human Factors in Computing Systems · doi:10.1145/3491102.3502030 · Cited by 345
Challenges in building intelligent open-domain dialog systems
Huang, M., Zhu, X., & Gao, J. (2020) · ACM Transactions on Information Systems · doi:10.1145/3383123 · Cited by 288
Gender and Representation Bias in GPT-3 Generated Stories
Lucy, L., & Bamman, D. (2021) · Proceedings of the Third Workshop on Narrative Understanding · doi:10.18653/v1/2021.nuse-1.5 · Cited by 287
Self-Diagnosis and Self-Debiasing: A Proposal for Reducing Corpus-Based Bias in NLP
Schick, T., Udupa, S., & Schütze, H. (2021) · Transactions of the Association for Computational Linguistics · doi:10.1162/tacl_a_00434 · Cited by 227
BBQ: A hand-built bias benchmark for question answering
Parrish, A., Chen, A., Nangia, N., Padmakumar, V., Phang, J., Thompson, J., Htut, P. M., & Bowman, S. (2022) · Findings of the Association for Computational Linguistics: ACL 2022 · doi:10.18653/v1/2022.findings-acl.165 · Cited by 186
Guo, W., & Caliskan, A. (2021) · Proceedings of the 2021 AAAI/ACM Conference on AI, Ethics, and Society · doi:10.1145/3461702.3462536 · Cited by 180
Stereotyping Norwegian Salmon: An Inventory of Pitfalls in Fairness Benchmark Datasets
Blodgett, S. L., Lopez, G., Olteanu, A., Sim, R., & Wallach, H. (2021) · Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers) · doi:10.18653/v1/2021.acl-long.81 · Cited by 147
Why Yellow Fever Isn't Flattering: A Case Against Racial Fetishes
ZHENG, R. (2016) · Journal of the American Philosophical Association · doi:10.1017/apa.2016.25 · Cited by 125
Barikeri, S., Lauscher, A., Vulić, I., & Glavaš, G. (2021) · Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers) · doi:10.18653/v1/2021.acl-long.151 · Cited by 121
Whose Opinions Do Language Models Reflect?
Santurkar, S., Durmus, E., Ladhak, F., Lee, C., Liang, P., & Hashimoto, T. (2023) · arXiv · doi:10.48550/arxiv.2303.17548 · Cited by 114
Self-citation“I’m sorry to hear that”: Finding new biases in language models with a holistic descriptor dataset
Smith, E. M., Hall, M., Kambadur, M., Presani, E., & Williams, A. (2022) · Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing · doi:10.18653/v1/2022.emnlp-main.625 · Cited by 97
Long Time No See! Open-Domain Conversation with Long-Term Persona Memory
Xu, X., Gou, Z., Wu, W., Niu, Z.-Y., Wu, H., Wang, H., & Wang, S. (2022) · arXiv · doi:10.48550/arxiv.2203.05797 · Cited by 65
Evidence for Hypodescent in Visual Semantic AI
Wolfe, R., Banaji, M. R., & Caliskan, A. (2022) · 2022 ACM Conference on Fairness Accountability and Transparency · doi:10.1145/3531146.3533185 · Cited by 19
Cited by
Bias and Fairness in Large Language Models: A Survey
Gallegos, I. O., Rossi, R. A., Barrow, J., Tanjim, M. M., Kim, S., Dernoncourt, F., Yu, T., Zhang, R., & Ahmed, N. K. (2024) · Computational Linguistics · doi:10.1162/coli_a_00524 · Cited by 1,245
AI generates covertly racist decisions about people based on their dialect
Hofmann, V., Kalluri, P. R., Jurafsky, D., & King, S. (2024) · Nature · doi:10.1038/s41586-024-07856-5 · Cited by 233
Self-citationSycophantic AI decreases prosocial intentions and promotes dependence
Cheng, M., Lee, C., Khadpe, P., Yu, S., Han, D., & Jurafsky, D. (2026) · Science · doi:10.1126/science.aec8352 · Cited by 185
Self-citationAccuracy and Political Bias of News Source Credibility Ratings by Large Language Models
Yang, K.-C., & Menczer, F. (2023) · arXiv · doi:10.48550/arxiv.2304.00228 · Cited by 115
Qin, H. X., Jin, S., Gao, Z., Fan, M., & Hui, P. (2024) · Proceedings of the CHI Conference on Human Factors in Computing Systems · doi:10.1145/3613904.3642105 · Cited by 82
Metadata
- Authors
- Cheng, M., Durmus, E., & Jurafsky, D.
- Venue
- Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
- Published
- 2023
- Pages
- 1504-1532
- OpenAlex
- W4385571220
- Cited by
- 115
- References
- 30
- FWCI
- 29.66
- Type
- conference paper
- Language
- en
- Open access
- hybrid
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.