More of the Same: Persistent Representational Harms Under Increased Representation

• URL: http://arxiv.org/abs/2503.00333v3
• Date: Fri, 31 Oct 2025 11:57:04 GMT
• Title: More of the Same: Persistent Representational Harms Under Increased Representation
• Authors: Jennifer Mickel, Maria De-Arteaga, Leqi Liu, Kevin Tian,
• Abstract summary: We develop an evaluation methodology for surfacing distribution-level group representational biases in generated text.<n>We show that, even though the gender distribution when models are prompted to generate biographies leads to a large representation of women, even representational biases persist in how different genders are represented.
• Score: 12.071592182704707
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/
• Abstract: To recognize and mitigate the harms of generative AI systems, it is crucial to consider whether and how different societal groups are represented by these systems. A critical gap emerges when naively measuring or improving who is represented, as this does not consider how people are represented. In this work, we develop GAS(P), an evaluation methodology for surfacing distribution-level group representational biases in generated text, tackling the setting where groups are unprompted (i.e., groups are not specified in the input to generative systems). We apply this novel methodology to investigate gendered representations in occupations across state-of-the-art large language models. We show that, even though the gender distribution when models are prompted to generate biographies leads to a large representation of women, even representational biases persist in how different genders are represented. Our evaluation methodology reveals that there are statistically significant distribution-level differences in the word choice used to describe biographies and personas of different genders across occupations, and we show that many of these differences are associated with representational harms and stereotypes. Our empirical findings caution that naively increasing (unprompted) representation may inadvertently proliferate representational biases, and our proposed evaluation methodology enables systematic and rigorous measurement of the problem.

Related papers

• BioPro: On Difference-Aware Gender Fairness for Vision-Language Models [50.40913324046528]
  Vision-Language Models (VLMs) inherit significant social biases from their training data, notably in gender representation.<n>We build upon recent advances in difference-aware fairness for text-only models to formalize the problem of difference-aware gender fairness for image captioning and text-to-image generation.<n>We propose BioPro, which aims to mitigate unwanted bias in neutral contexts while preserving valid distinctions in explicit ones.
  arXiv  Detail & Related papers  (2025-11-30T09:33:09Z)
• Gender Bias in Instruction-Guided Speech Synthesis Models [55.2480439325792]
  This study investigates the potential gender bias in how models interpret occupation-related prompts.<n>We explore whether these models exhibit tendencies to amplify gender stereotypes when interpreting such prompts.<n>Our experimental results reveal the model's tendency to exhibit gender bias for certain occupations.
  arXiv  Detail & Related papers  (2025-02-08T17:38:24Z)
• Blind Men and the Elephant: Diverse Perspectives on Gender Stereotypes in Benchmark Datasets [12.798832545154271]
  This paper examines the inconsistencies between intrinsic stereotype benchmarks.<n>Using StereoSet and CrowS-Pairs as case studies, we investigated how data distribution affects benchmark results.
  arXiv  Detail & Related papers  (2025-01-02T09:40:31Z)
• Evaluating Gender Bias in Large Language Models [0.8636148452563583]
  The study examines the extent to which Large Language Models (LLMs) exhibit gender bias in pronoun selection in occupational contexts. The jobs considered include a range of occupations, from those with a significant male presence to those with a notable female concentration. The results show a positive correlation between the models' pronoun choices and the gender distribution present in U.S. labor force data.
  arXiv  Detail & Related papers  (2024-11-14T22:23:13Z)
• 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)
• Generalizing Fairness to Generative Language Models via Reformulation of Non-discrimination Criteria [4.738231680800414]
  This paper studies how to uncover and quantify the presence of gender biases in generative language models. We derive generative AI analogues of three well-known non-discrimination criteria from classification, namely independence, separation and sufficiency. Our results address the presence of occupational gender bias within such conversational language models.
  arXiv  Detail & Related papers  (2024-03-13T14:19:08Z)
• The Male CEO and the Female Assistant: Evaluation and Mitigation of Gender Biases in Text-To-Image Generation of Dual Subjects [58.27353205269664]
  We propose the Paired Stereotype Test (PST) framework, which queries T2I models to depict two individuals assigned with male-stereotyped and female-stereotyped social identities. PST queries T2I models to depict two individuals assigned with male-stereotyped and female-stereotyped social identities. Using PST, we evaluate two aspects of gender biases -- the well-known bias in gendered occupation and a novel aspect: bias in organizational power.
  arXiv  Detail & Related papers  (2024-02-16T21:32:27Z)
• Stable Diffusion Exposed: Gender Bias from Prompt to Image [25.702257177921048]
  This paper introduces an evaluation protocol that analyzes the impact of gender indicators at every step of the generation process on Stable Diffusion images. Our findings include the existence of differences in the depiction of objects, such as instruments tailored for specific genders, and shifts in overall layouts.
  arXiv  Detail & Related papers  (2023-12-05T10:12:59Z)
• VisoGender: A dataset for benchmarking gender bias in image-text pronoun resolution [80.57383975987676]
  VisoGender is a novel dataset for benchmarking gender bias in vision-language models. We focus on occupation-related biases within a hegemonic system of binary gender, inspired by Winograd and Winogender schemas. We benchmark several state-of-the-art vision-language models and find that they demonstrate bias in resolving binary gender in complex scenes.
  arXiv  Detail & Related papers  (2023-06-21T17:59:51Z)
• Stereotypes and Smut: The (Mis)representation of Non-cisgender Identities by Text-to-Image Models [6.92043136971035]
  We investigate how multimodal models handle diverse gender identities. We find certain non-cisgender identities are consistently (mis)represented as less human, more stereotyped and more sexualised. These improvements could pave the way for a future where change is led by the affected community.
  arXiv  Detail & Related papers  (2023-05-26T16:28:49Z)
• Fairness in AI Systems: Mitigating gender bias from language-vision models [0.913755431537592]
  We study the extent of the impact of gender bias in existing datasets. We propose a methodology to mitigate its impact in caption based language vision models.
  arXiv  Detail & Related papers  (2023-05-03T04:33:44Z)
• Stable Bias: Analyzing Societal Representations in Diffusion Models [72.27121528451528]
  We propose a new method for exploring the social biases in Text-to-Image (TTI) systems. Our approach relies on characterizing the variation in generated images triggered by enumerating gender and ethnicity markers in the prompts. We leverage this method to analyze images generated by 3 popular TTI systems and find that while all of their outputs show correlations with US labor demographics, they also consistently under-represent marginalized identities to different extents.
  arXiv  Detail & Related papers  (2023-03-20T19:32:49Z)
• Auditing Gender Presentation Differences in Text-to-Image Models [54.16959473093973]
  We study how gender is presented differently in text-to-image models. By probing gender indicators in the input text, we quantify the frequency differences of presentation-centric attributes. We propose an automatic method to estimate such differences.
  arXiv  Detail & Related papers  (2023-02-07T18:52:22Z)
• Easily Accessible Text-to-Image Generation Amplifies Demographic Stereotypes at Large Scale [61.555788332182395]
  We investigate the potential for machine learning models to amplify dangerous and complex stereotypes. We find a broad range of ordinary prompts produce stereotypes, including prompts simply mentioning traits, descriptors, occupations, or objects.
  arXiv  Detail & Related papers  (2022-11-07T18:31:07Z)
• Are Commercial Face Detection Models as Biased as Academic Models? [64.71318433419636]
  We compare academic and commercial face detection systems, specifically examining robustness to noise. We find that state-of-the-art academic face detection models exhibit demographic disparities in their noise robustness. We conclude that commercial models are always as biased or more biased than an academic model.
  arXiv  Detail & Related papers  (2022-01-25T02:21:42Z)
• Social Norm Bias: Residual Harms of Fairness-Aware Algorithms [21.50551404445654]
  Social Norm Bias (SNoB) is a subtle but consequential type of discrimination that may be exhibited by automated decision-making systems. We quantify SNoB by measuring how an algorithm's predictions are associated with conformity to gender norms. We show that post-processing interventions do not mitigate this type of bias at all.
  arXiv  Detail & Related papers  (2021-08-25T05:54:56Z)
• Towards Understanding and Mitigating Social Biases in Language Models [107.82654101403264]
  Large-scale pretrained language models (LMs) can be potentially dangerous in manifesting undesirable representational biases. We propose steps towards mitigating social biases during text generation. Our empirical results and human evaluation demonstrate effectiveness in mitigating bias while retaining crucial contextual information.
  arXiv  Detail & Related papers  (2021-06-24T17:52:43Z)

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.