mlsanityData to 14 Oct 2026

Causally Testing Gender Bias in LLMs: A Case Study on Occupational Bias

arXiv 2212.10678 ↗First posted 20 Dec 2022cs.CL cs.LGcited by 28

Not on the map. Why?

Abstract (arXiv)

Generated texts from large language models (LLMs) have been shown to exhibit a variety of harmful, human-like biases against various demographics. These findings motivate research efforts aiming to understand and measure such effects. This paper introduces a causal formulation for bias measurement in generative language models. Based on this theoretical foundation, we outline a list of desiderata for designing robust bias benchmarks. We then propose a benchmark called OccuGender, with a bias-measuring procedure to investigate occupational gender bias. We test several state-of-the-art open-source LLMs on OccuGender, including Llama, Mistral, and their instruction-tuned versions. The results show that these models exhibit substantial occupational gender bias. Lastly, we discuss prompting strategies for bias mitigation and an extension of our causal formulation to illustrate the generalizability of our framework. Our code and data https://github.com/chenyuen0103/gender-bias.

Key facts. Cited by 28 papers in the map (data through 14 Oct 2026). Builds on 42 of its 77 references that are in the map. First posted on arXiv 20 Dec 2022.

Builds on

42 of 77 references.

References · 42

  1. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,399
  2. 2020Language Models are Few-Shot Learnerscited by 32,419
  3. 2019Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformercited by 14,599
  4. 2022PaLM: Scaling Language Modeling with Pathwayscited by 5,326
  5. 2021Finetuned Language Models Are Zero-Shot Learnerscited by 3,780
  6. 2019A Survey on Bias and Fairness in Machine Learningcited by 1,917
  7. 2016Man is to Computer Programmer as Woman is to Homemaker? Debiasing Word Embeddingscited by 1,312
  8. 2020Making Pre-trained Language Models Better Few-shot Learnerscited by 1,256
  9. 2020RealToxicityPrompts: Evaluating Neural Toxic Degeneration in Language Modelscited by 1,255
  10. 2020Beyond Accuracy: Behavioral Testing of NLP models with CheckListcited by 1,006
  11. 2021Ethical and social risks of harm from Language Modelscited by 966
  12. 2020Language (Technology) is Power: A Critical Survey of "Bias" in NLPcited by 956
30 more
  1. 2020StereoSet: Measuring stereotypical bias in pretrained language modelscited by 924
  2. 2020CrowS-Pairs: A Challenge Dataset for Measuring Social Biases in Masked Language Modelscited by 710
  3. 2018Mitigating Unwanted Biases with Adversarial Learningcited by 663
  4. 2018Gender Bias in Coreference Resolution: Evaluation and Debiasing Methodscited by 659
  5. 2019The Woman Worked as a Babysitter: On Biases in Language Generationcited by 541
  6. 2015Causal inference using invariant prediction: identification and confidence intervalscited by 536
  7. 2021Retiring Adult: New Datasets for Fair Machine Learningcited by 449
  8. 2018Gender Bias in Coreference Resolutioncited by 442
  9. 2021BOLD: Dataset and Metrics for Measuring Biases in Open-Ended Language Generationcited by 412
  10. 2021Persistent Anti-Muslim Bias in Large Language Modelscited by 377
  11. 2021Self-Diagnosis and Self-Debiasing: A Proposal for Reducing Corpus-Based Bias in NLPcited by 337
  12. 2021Towards Understanding and Mitigating Social Biases in Language Modelscited by 296
  13. 2020Measuring and Reducing Gendered Correlations in Pre-trained Modelscited by 224
  14. 2021Towards Causal Representation Learningcited by 209
  15. 2019Gender Bias in Contextualized Word Embeddingscited by 203
  16. 2019Black is to Criminal as Caucasian is to Police: Detecting and Removing Multiclass Bias in Word Embeddingscited by 147
  17. 2017Fair Inference On Outcomescited by 144
  18. 2021Bias Out-of-the-Box: An Empirical Analysis of Intersectional Occupational Biases in Popular Generative Language Modelscited by 141
  19. 2021Harms of Gender Exclusivity and Challenges in Non-Binary Representation in Language Technologiescited by 128
  20. 2022GrIPS: Gradient-free, Edit-based Instruction Search for Prompting Large Language Modelscited by 127
  21. 2021A Recipe For Arbitrary Text Style Transfer with Large Language Modelscited by 107
  22. 2021A Survey on Gender Bias in Natural Language Processingcited by 89
  23. 2020Investigating Bias and Fairness in Facial Expression Recognitioncited by 87
  24. 2021AutoDebias: Learning to Debias for Recommendationcited by 68
  25. 2022A Causal Framework to Quantify the Robustness of Mathematical Reasoning with Language Modelscited by 68
  26. 2021PASS: An ImageNet replacement for self-supervised pretraining without humanscited by 51
  27. 2022A Causal Lens for Controllable Text Generationcited by 45
  28. 2020Mitigating Gender Bias Amplification in Distribution by Posterior Regularizationcited by 27
  29. 2022Challenges in Measuring Bias via Open-Ended Language Generationcited by 24
  30. 2020LOGAN: Local Group Bias Detection by Clusteringcited by 9

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2023: 9; 2024: 6; 2025: 7; 2026: 6.

28 papers.

Sort:

By area:

  1. 2023Bias and Fairness in Large Language Models: A Surveycited by 639
  2. 2023A Survey on Fairness in Large Language Modelscited by 95
  3. 2023Language Model Behavior: A Comprehensive Surveycited by 92
  4. 2024Self-Debiasing Large Language Models: Zero-Shot Recognition and Reduction of Stereotypescited by 59
  5. 2024Social Bias Evaluation for Large Language Models Requires Prompt Variationscited by 40
  6. 2024Dialect prejudice predicts AI decisions about people's character, employability, and criminalitycited by 37
  7. 2023How Trustworthy are Open-Source LLMs? An Assessment under Malicious Demonstrations Shows their Vulnerabilitiescited by 33
  8. 2023Towards Trustworthy and Aligned Machine Learning: A Data-centric Survey with Causality Perspectivescited by 30
  9. 2023Unveiling Gender Bias in Terms of Profession Across LLMs: Analyzing and Addressing Sociological Implicationscited by 29
  10. 2023Tackling Bias in Pre-trained Language Models: Current Trends and Under-represented Societiescited by 11
  11. 2024Potential and Challenges of Model Editing for Social Debiasingcited by 10
  12. 2023Who's Thinking? A Push for Human-Centered Evaluation of LLMs using the XAI Playbookcited by 5
16 more
  1. 2023Scratch Copilot Evaluation: Assessing AI-Assisted Creative Coding for Familiescited by 5
  2. 2025Beyond Prompt-Induced Lies: Investigating LLM Deception on Benign Promptscited by 5
  3. 2024Interpreting Bias in Large Language Models: A Feature-Based Approachcited by 4
  4. 2024Improving Linguistic Diversity of Large Language Models with Possibility Exploration Fine-Tuningcited by 4
  5. 2025Bias after Prompting: Persistent Discrimination in Large Language Modelscited by 3
  6. 2025Generating the Modal Worker: A Cross-Model Audit of Race and Gender in LLM-Generated Personas Across 41 Occupationscited by 3
  7. 2025Causality Is Key to Understand and Balance Multiple Goals in Trustworthy ML and Foundation Modelscited by 2
  8. 2025Dropouts in Confidence: Moral Uncertainty in Human-LLM Alignmentcited by 2
  9. 2026VEAT Quantifies Implicit Associations in Text-to-Video Generator Sora and Reveals Challenges in Bias Mitigationcited by 2
  10. 2025Gender Encoding Patterns in Pretrained Language Model Representationscited by 1
  11. 2026Personality Shapes Gender Bias in Persona-Conditioned LLM Narratives Across English and Hindi: An Empirical Investigationcited by 1
  12. 2025Bridging Fairness and Explainability: Can Input-Based Explanations Promote Fairness in Hate Speech Detection?cited by 0
  13. 2026Evaluating the Usage of African-American Vernacular English in Large Language Modelscited by 0
  14. 2026GKnow: Measuring the Entanglement of Gender Bias and Factual Gendercited by 0
  15. 2026Who Should Be Generated? Justifying Demographic Targets in Open-Ended Generationcited by 0
  16. 2026LLMs Mirror Country-Specific Gender Patterns If Asked, but Skew Male When Generating Media in Local Languagescited by 0