CoChat

Persistent Anti-Muslim Bias in Large Language Models

Open access
Abid, Abubakar, Farooqi, Maheen, Zou, James · 2021 · Proceedings of the 2021 AAAI/ACM Conference on AI, Ethics, and Society

The DOI matches its Crossref record; 13 of the 13 references it cites were found in a registry or an index.

Abstract

It has been observed that large-scale language models capture undesirable societal biases, e.g. relating to race and gender; yet religious bias has been relatively unexplored. We demonstrate that GPT-3, a state-of-the-art contextual language model, captures persistent Muslim-violence bias. We probe GPT-3 in various ways, including prompt completion, analogical reasoning, and story generation, to understand this anti-Muslim bias, demonstrating that it appears consistently and creatively in different uses of the model and that it is severe even compared to biases about other religious groups. For instance, Muslim is analogized to terrorist in 23% of test cases, while Jewish is mapped to its most common stereotype, money, in 5% of test cases. We quantify the positive distraction needed to overcome this bias with adversarial text prompts, and find that use of the most positive 6 adjectives reduces violent completions for Muslims from 66% to 20%, but which is still higher than for other religious groups.

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.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1145/3461702.3462624

Cites

13 of 13 references verified0 citation errors · 0 not checkable

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.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

1 more claim about this paper

Create a free account to see every claim other papers make about this one.

See all claims — free account

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

  • Attention Is All You Need

    Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., Kaiser, L., & Polosukhin, I. (2017) · arXiv · doi:10.48550/arxiv.1706.03762 · Cited by 120,404 · cited 1×

    “These languagemodels have evolved, from learning individual word vectors with single-layer models [8], to more complex language generation architectures such as recurrent neural networks [5] and most recently transformers [4, 6, 14].”
    Background
  • Efficient Estimation of Word Representations in Vector Space

    Mikolov, T., Chen, K., Corrado, G., & Dean, J. (2013) · arXiv · doi:10.48550/arxiv.1301.3781 · Cited by 26,568 · cited 1×

    “These languagemodels have evolved, from learning individual word vectors with single-layer models [8], to more complex language generation architectures such as recurrent neural networks [5] and most recently transformers [4, 6, 14].”
    Background
  • Language Models are Few-Shot Learners

    Brown, T. B., Mann, B., Ryder, N., Subbiah, M., Kaplan, J., Dhariwal, P., Neelakantan, A., Shyam, P., Sastry, G., Askell, A., et al. (2020) · arXiv · doi:10.48550/arxiv.2005.14165 · Cited by 23,372 · cited 4×

    “These languagemodels have evolved, from learning individual word vectors with single-layer models [8], to more complex language generation architectures such as recurrent neural networks [5] and most recently transformers [4, 6, 14].”
    BackgroundMethodInfluential
  • Self-Critical Sequence Training for Image Captioning

    Rennie, S. J., Marcheret, E., Mroueh, Y., Ross, J., & Goel, V. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.131 · Cited by 2,112 · cited 1×

    “These descriptions are generated by passing images through a pretrained image captioning model [11], and fine-tuning GPT-3 to transform the short captions into long descriptions, as shown in Fig.”
    Method
  • Reformer: The Efficient Transformer

    Kitaev, N., Kaiser, Ł., & Levskaya, A. (2020) · arXiv · doi:10.48550/arxiv.2001.04451 · Cited by 1,435 · cited 1×

    “These languagemodels have evolved, from learning individual word vectors with single-layer models [8], to more complex language generation architectures such as recurrent neural networks [5] and most recently transformers [4, 6, 14].”
    Background
  • Semi-supervised Sequence Learning

    Dai, A. M., & Le, Q. V. (2015) · arXiv · doi:10.48550/arxiv.1511.01432 · Cited by 994

  • 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 2×

    “As a result, even though the various language models have different architectures, since they are trained on similar corpora of text, they often learn similar biases [9].”
    Background
  • Universal Adversarial Triggers for Attacking and Analyzing NLP

    Wallace, E., Feng, S., Kandpal, N., Gardner, M., & Singh, S. (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/d19-1221 · Cited by 642

  • Identifying and Reducing Gender Bias in Word-Level Language Models

    Bordia, S., & Bowman, S. R. (2019) · Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Student Research Workshop · doi:10.18653/v1/n19-3002 · Cited by 219 · cited 1×

    “Previous work has explored the biases that are ultimately learned by language models, with the primary focus being racial and gender stereotypes [2, 3, 7, 12].”
    Background
  • The Woman Worked as a Babysitter: On Biases in Language GenerationPreprint

    Sheng, E., Chang, K.-W., Natarajan, P., & Peng, N. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1909.01326 · Cited by 205 · cited 2×

    “In [12], the authors made similar observations about language generation models.”
    Background

Show all 13 references

Research works printed but not linked: 6

  1. 7. T. Bolukbasi, K.-W. Chang, J. Y. Zou, V. Saligrama, A. T. Kalai, Advances in neural information processing systems 29 , 4349 (2016).
  2. 9. S. Bordia, S. R. Bowman, arXiv preprint arXiv:1904.03035 (2019). · arXiv
  3. 2. A. M. Dai, Q. V. Le, Advances in neural information processing systems 28 , 3079 (2015).
  4. 10. K. Lu, P. Mardziel, F. Wu, P. Amancharla, A. Datta, Logic, Language, and Security (Springer, 2020), pp. 189-202.
  5. 3. A. Vaswani, et al. , Advances in neural information processing systems (2017), pp. 59986008.
  6. 14. E. Wallace, S. Feng, N. Kandpal, M. Gardner, S. Singh, arXiv preprint arXiv:1908.07125 (2019). · arXiv

Cited by

Items 401–500 of 986

Metadata

Authors
Abid, A., Farooqi, M., & Zou, J.
Affiliations
  • Stanford University (US)
  • McMaster University (CA)
Venue
Proceedings of the 2021 AAAI/ACM Conference on AI, Ethics, and Society
Publisher
ACM
Published
2021-07-21
Pages
298-306
ISSN
3065-8365 (ISSN-L)
OpenAlex
W3184144760
Cited by
476
References
3
Citations by year
2026: 81 · 2025: 124 · 2024: 105 · 2023: 97 · 2022: 36 · 2021: 33
FWCI
61.36
Type
conference paper
Language
en
Open access
Locations
  • Proceedings of the AAAI/ACM Conference on AI Ethics and Society (publishedVersion)
  • arXiv (Cornell University) (submittedVersion, cc-by)
  • arXiv (Cornell University) (cc-by)
Topics
  • Physical Sciences › Computer Science › Artificial Intelligence › Ethics and Social Impacts of AI
  • Physical Sciences › Computer Science › Artificial Intelligence › Hate Speech and Cyberbullying Detection
  • Physical Sciences › Computer Science › Artificial Intelligence › Topic Modeling
Keywords
Islamophobia, religious bias, large language models, GPT-3, analogical reasoning, adversarial prompts, story generation, bias mitigation

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.