Measuring Harmful Sentence Completion in Language Models for LGBTQIA+ Individuals
Open accessThe DOI matches its Crossref record; 13 of the 13 references it cites were found in a registry or an index.
Abstract
Warning: This paper contains examples of language that some people may find offensive or upsetting.Current language technology is ubiquitous and directly influences individuals' lives worldwide.Given the recent trend in AI on training and constantly releasing new and powerful large language models (LLMs), there is a need to assess their biases and potential concrete consequences.While some studies have highlighted the shortcomings of these models, there is only little on the negative impact of LLMs on LGBTQIA+ individuals.In this paper, we investigated a state-of-the-art template-based approach for measuring the harmfulness of English LLMs sentence completion when the subjects belong to the LGBTQIA+ community.Our findings show that, on average, the most likely LLM-generated completion is an identity attack 13% of the time.Our results raise serious concerns about the applicability of these models in production environments.
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
BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
Devlin, J., Chang, M.-W., Lee, K., & Toutanova, K. (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/n19-1423 · Cited by 54,325
Transformers: State-of-the-Art Natural Language Processing
Wolf, T., Debut, L., Sanh, V., Chaumond, J., Delangue, C., Moi, A., Cistac, P., Rault, T., Louf, R., Funtowicz, M., et al. (2020) · Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing: System Demonstrations · doi:10.18653/v1/2020.emnlp-demos.6 · Cited by 8,741
Language (Technology) is Power: A Critical Survey of “Bias” in NLP
Blodgett, S. L., Barocas, S., Daumé III, H., & Wallach, H. (2020) · Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics · doi:10.18653/v1/2020.acl-main.485 · Cited by 1,050 · cited 1×
“Several works in NLP have investigated issues arising from unfair language technology (Hovy and Spruit, 2016; Blodgett et al., 2020; Bianchi and Hovy, 2021), such as stereotypically biased and polarized embeddings (e.g., Bolukbasi et al., 2016; Gonen and Goldberg, 2019; Lauscher et al., 2020;…”
BackgroundStereoSet: 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
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 · cited 1×
“Recent work has proposed the crowdsourcing collection of stereotypes also related to gender identity and sexual orientation (Nangia et al., 2020; Nadeem et al., 2021).”
BackgroundThe Woman Worked as a Babysitter: On Biases in Language Generation
Sheng, E., Chang, K.-W., Natarajan, P., & Peng, N. (2019) · Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP) · doi:10.18653/v1/d19-1339 · Cited by 537 · cited 1×
“…2016; Gonen and Goldberg, 2019; Lauscher et al., 2020; Bianchi et al., 2021), and unfairly biased task-specific models, e.g., for natural language inference (e.g., Dev et al., 2020), natural language generation (e.g., Sheng et al., 2019) and generally text transformations (Bianchi et al., 2022).”
BackgroundMitigating Gender Bias in Natural Language Processing: Literature Review
Sun, T., Gaut, A., Tang, S., Huang, Y., ElSherief, M., Zhao, J., Mirza, D., Belding, E., Chang, K.-W., & Wang, W. Y. (2019) · Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics · doi:10.18653/v1/p19-1159 · Cited by 492
Gonen, H., & Goldberg, Y. (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/n19-1061 · Cited by 454 · cited 2×
“…Spruit, 2016; Blodgett et al., 2020; Bianchi and Hovy, 2021), such as stereotypically biased and polarized embeddings (e.g., Bolukbasi et al., 2016; Gonen and Goldberg, 2019; Lauscher et al., 2020; Bianchi et al., 2021), and unfairly biased task-specific models, e.g., for natural language…”
BackgroundCrowS-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
The social impact of natural language processing
Hovy, D., & Spruit, S. L. (2016) · Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers) · doi:10.18653/v1/p16-2096 · Cited by 329
Self-citation
Cited by
Biases in Large Language Models: Origins, Inventory, and Discussion
Navigli, R., Conia, S., & Ross, B. (2023) · Journal of Data and Information Quality · doi:10.1145/3597307 · Cited by 364
The life cycle of large language models in education: A framework for understanding sources of bias
Lee, J., Hicke, Y., Yu, R., Brooks, C., & Kizilcec, R. F. (2024) · British Journal of Educational Technology · doi:10.1111/bjet.13505 · Cited by 119
“Examples of lexicon-based metrics include HONEST (Nozza et al., 2022), which measures harmful words in generated text, and BOLD (Dhamala et al., 2021), which measures psycho-linguistic norms by assigning affective values (eg, 14678535 , 0 , D o w n l o a d e d fr o m h tt p s :// b e r a - j ou r n…”
BackgroundLanguage Model Behavior: A Comprehensive Survey
Chang, T. A., & Bergen, B. K. (2024) · Computational Linguistics · doi:10.1162/coli_a_00492 · Cited by 105
“13% of the time in masked models (350M parameters), up to 87% for some identity groups (Nozza et al. 2022).”
BackgroundHaving Beer after Prayer? Measuring Cultural Bias in Large Language Models
Naous, T., Ryan, M. J., Ritter, A., & Xu, W. (2024) · Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · doi:10.18653/v1/2024.acl-long.862 · Cited by 96
XSTest: A Test Suite for Identifying Exaggerated Safety Behaviours in Large Language Models
Röttger, P., Kirk, H., Vidgen, B., Attanasio, G., Bianchi, F., & Hovy, D. (2024) · Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers) · doi:10.18653/v1/2024.naacl-long.301 · Cited by 77
Self-citation
Metadata
- Authors
- Nozza, D., Bianchi, F., Lauscher, A., & Hovy, D.
- Venue
- Proceedings of the Second Workshop on Language Technology for Equality, Diversity and Inclusion
- Published
- 2022
- Pages
- 26-34
- OpenAlex
- W4285210581
- Cited by
- 44
- References
- 13
- FWCI
- 9.76
- Type
- conference paper
- Language
- en
- Open access
- gold
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
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Measuring Harmful Sentence Completion in Language Models for LGBTQIA+ Individuals. https://app.cochat.ai/papers/10.18653/v1/2022.ltedi-1.4
- The paper
- Nozza, D., Bianchi, F., Lauscher, A., & Hovy, D. (2022). Measuring Harmful Sentence Completion in Language Models for LGBTQIA+ Individuals. Proceedings of the Second Workshop on Language Technology for Equality, Diversity and Inclusion. https://doi.org/10.18653/v1/2022.ltedi-1.4
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.