Measuring Bias or Measuring the Task: Understanding the Brittle Nature of LLM Gender Biases

• URL: http://arxiv.org/abs/2509.04373v3
• Date: Wed, 10 Sep 2025 06:08:26 GMT
• Title: Measuring Bias or Measuring the Task: Understanding the Brittle Nature of LLM Gender Biases
• Authors: Bufan Gao, Elisa Kreiss,
• Abstract summary: This paper examines how signaling the evaluative purpose of a task impacts measured gender bias in LLMs.<n>We find that prompts that more clearly align with (gender bias) evaluation framing elicit distinct gender output distributions.
• Score: 2.9803250365852443
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: As LLMs are increasingly applied in socially impactful settings, concerns about gender bias have prompted growing efforts both to measure and mitigate such bias. These efforts often rely on evaluation tasks that differ from natural language distributions, as they typically involve carefully constructed task prompts that overtly or covertly signal the presence of gender bias-related content. In this paper, we examine how signaling the evaluative purpose of a task impacts measured gender bias in LLMs. Concretely, we test models under prompt conditions that (1) make the testing context salient, and (2) make gender-focused content salient. We then assess prompt sensitivity across four task formats with both token-probability and discrete-choice metrics. We find that prompts that more clearly align with (gender bias) evaluation framing elicit distinct gender output distributions compared to less evaluation-framed prompts. Discrete-choice metrics further tend to amplify bias relative to probabilistic measures. These findings do not only highlight the brittleness of LLM gender bias evaluations but open a new puzzle for the NLP benchmarking and development community: To what extent can well-controlled testing designs trigger LLM "testing mode" performance, and what does this mean for the ecological validity of future benchmarks.

Related papers

• The Confidence Trap: Gender Bias and Predictive Certainty in LLMs [5.926203312586108]
  The research investigates probability confidence calibration in contexts involving gendered pronoun resolution.<n>The goal is to evaluate if calibration metrics based on predicted confidence scores effectively capture fairness-related disparities in Large Language Models.
  arXiv  Detail & Related papers  (2026-01-12T18:38:05Z)
• Bias in Gender Bias Benchmarks: How Spurious Features Distort Evaluation [116.86965910589775]
  We show that even minimal perturbations, such as masking just 10% of objects or weakly blurring backgrounds, can dramatically alter bias scores.<n>This suggests that current bias evaluations reflect model responses to spurious features rather than gender bias.
  arXiv  Detail & Related papers  (2025-09-09T11:14:11Z)
• Towards Fair Rankings: Leveraging LLMs for Gender Bias Detection and Measurement [6.92803536773427]
  Social biases in Natural Language Processing (NLP) and Information Retrieval (IR) systems are an ongoing challenge.<n>We aim to address this issue by leveraging Large Language Models (LLMs) to detect and measure gender bias in passage ranking.<n>We introduce a novel gender fairness metric, named Class-wise Weighted Exposure (CWEx), aiming to address existing limitations.
  arXiv  Detail & Related papers  (2025-06-27T16:39:12Z)
• EMO-Debias: Benchmarking Gender Debiasing Techniques in Multi-Label Speech Emotion Recognition [49.27067541740956]
  EMO-Debias is a large-scale comparison of 13 debiasing methods applied to multi-label SER.<n>Our study encompasses techniques from pre-processing, regularization, adversarial learning, biased learners, and distributionally robust optimization.<n>Our analysis quantifies the trade-offs between fairness and accuracy, identifying which approaches consistently reduce gender performance gaps.
  arXiv  Detail & Related papers  (2025-06-05T05:48:31Z)
• The Root Shapes the Fruit: On the Persistence of Gender-Exclusive Harms in Aligned Language Models [91.86718720024825]
  We center transgender, nonbinary, and other gender-diverse identities to investigate how alignment procedures interact with pre-existing gender-diverse bias.<n>Our findings reveal that DPO-aligned models are particularly sensitive to supervised finetuning.<n>We conclude with recommendations tailored to DPO and broader alignment practices.
  arXiv  Detail & Related papers  (2024-11-06T06:50:50Z)
• GenderCARE: A Comprehensive Framework for Assessing and Reducing Gender Bias in Large Language Models [73.23743278545321]
  Large language models (LLMs) have exhibited remarkable capabilities in natural language generation, but have also been observed to magnify societal biases.<n>GenderCARE is a comprehensive framework that encompasses innovative Criteria, bias Assessment, Reduction techniques, and Evaluation metrics.
  arXiv  Detail & Related papers  (2024-08-22T15:35:46Z)
• The African Woman is Rhythmic and Soulful: An Investigation of Implicit Biases in LLM Open-ended Text Generation [3.9945212716333063]
  Implicit biases are significant because they influence the decisions made by Large Language Models (LLMs) Traditionally, explicit bias tests or embedding-based methods are employed to detect bias, but these approaches can overlook more nuanced, implicit forms of bias. We introduce two novel psychological-inspired methodologies to reveal and measure implicit biases through prompt-based and decision-making tasks.
  arXiv  Detail & Related papers  (2024-07-01T13:21:33Z)
• GenderBias-\emph{VL}: Benchmarking Gender Bias in Vision Language Models via Counterfactual Probing [72.0343083866144]
  This paper introduces the GenderBias-emphVL benchmark to evaluate occupation-related gender bias in Large Vision-Language Models. Using our benchmark, we extensively evaluate 15 commonly used open-source LVLMs and state-of-the-art commercial APIs. Our findings reveal widespread gender biases in existing LVLMs.
  arXiv  Detail & Related papers  (2024-06-30T05:55:15Z)
• Probing Explicit and Implicit Gender Bias through LLM Conditional Text Generation [64.79319733514266]
  Large Language Models (LLMs) can generate biased and toxic responses. We propose a conditional text generation mechanism without the need for predefined gender phrases and stereotypes.
  arXiv  Detail & Related papers  (2023-11-01T05:31:46Z)
• Causally Testing Gender Bias in LLMs: A Case Study on Occupational Bias [33.99768156365231]
  We introduce a causal formulation for bias measurement in generative language models.<n>We propose a benchmark called OccuGender, with a bias-measuring procedure to investigate occupational gender bias.<n>The results show that these models exhibit substantial occupational gender bias.
  arXiv  Detail & Related papers  (2022-12-20T22:41:24Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.