mlsanityData to 7 Oct 2026

SocialCounterfactuals: Probing and Mitigating Intersectional Social Biases in Vision-Language Models with Counterfactual Examples

arXiv 2312.00825 ↗First posted 30 Nov 2023cs.CV cs.AIcited by 53

On the map: Multimodal LLMs / vision-language models → Social bias and fairness in VLMs

also in 1 more
Abstract (arXiv)

While vision-language models (VLMs) have achieved remarkable performance improvements recently, there is growing evidence that these models also posses harmful biases with respect to social attributes such as gender and race. Prior studies have primarily focused on probing such bias attributes individually while ignoring biases associated with intersections between social attributes. This could be due to the difficulty of collecting an exhaustive set of image-text pairs for various combinations of social attributes. To address this challenge, we employ text-to-image diffusion models to produce counterfactual examples for probing intersectional social biases at scale. Our approach utilizes Stable Diffusion with cross attention control to produce sets of counterfactual image-text pairs that are highly similar in their depiction of a subject (e.g., a given occupation) while differing only in their depiction of intersectional social attributes (e.g., race & gender). Through our over-generate-then-filter methodology, we produce SocialCounterfactuals, a high-quality dataset containing 171k image-text pairs for probing intersectional biases related to gender, race, and physical characteristics. We conduct extensive experiments to demonstrate the usefulness of our generated dataset for probing and mitigating intersectional social biases in state-of-the-art VLMs.

Key facts. Cited by 53 papers in the map (data through 7 Oct 2026). Placed in Social bias and fairness in VLMs, one of 16 directions in Multimodal LLMs / vision-language models. Builds on 53 of its 66 references that are in the map. First posted on arXiv 30 Nov 2023.

Builds on

53 of 66 references.

Closest ancestors · 6

Same direction, newest first

  1. 2023VisoGender: A dataset for benchmarking gender bias in image-text pronoun resolution21 Jun
  2. 2023Balancing the Picture: Debiasing Vision-Language Datasets with Synthetic Contrast Sets24 May
  3. 2023DeAR: Debiasing Vision-Language Models with Additive Residuals18 Mar
  4. 2023Debiasing Vision-Language Models via Biased Prompts31 Jan
  5. 2023Vision-Language Models Performing Zero-Shot Tasks Exhibit Gender-based Disparities26 Jan
  6. 2022A Prompt Array Keeps the Bias Away: Debiasing Vision-Language Models with Adversarial Learning22 Mar

Other references · 47

  1. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,456
  2. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,694
  3. 2022InstructPix2Pix: Learning to Follow Image Editing Instructionscited by 2,591
  4. 2022Prompt-to-Prompt Image Editing with Cross Attention Controlcited by 2,409
  5. 2020Shortcut Learning in Deep Neural Networkscited by 1,928
  6. 2022Reproducible scaling laws for contrastive language-image learningcited by 1,163
  7. 2020StereoSet: Measuring stereotypical bias in pretrained language modelscited by 911
  8. 2020CrowS-Pairs: A Challenge Dataset for Measuring Social Biases in Masked Language Modelscited by 703
  9. 2018Gender Bias in Coreference Resolution: Evaluation and Debiasing Methodscited by 657
  10. 2021FLAVA: A Foundational Language And Vision Alignment Modelcited by 646
  11. 2018Gender Bias in Coreference Resolutioncited by 438
  12. 2021SLIP: Self-supervision meets Language-Image Pre-trainingcited by 434
35 more
  1. 2019Learning the Difference that Makes a Difference with Counterfactually-Augmented Datacited by 428
  2. 2017Preventing Fairness Gerrymandering: Auditing and Learning for Subgroup Fairnesscited by 381
  3. 2020Noise or Signal: The Role of Image Backgrounds in Object Recognitioncited by 342
  4. 2021Multimodal datasets: misogyny, pornography, and malignant stereotypescited by 326
  5. 2019Towards Fairness in Visual Recognition: Effective Strategies for Bias Mitigationcited by 252
  6. 2022DALL-Eval: Probing the Reasoning Skills and Social Biases of Text-to-Image Generation Modelscited by 230
  7. 2021Polyjuice: Generating Counterfactuals for Explaining, Evaluating, and Improving Modelscited by 219
  8. 2023Improving CLIP Training with Language Rewritescited by 216
  9. 2022"I'm sorry to hear that": Finding New Biases in Language Models with a Holistic Descriptor Datasetcited by 160
  10. 2023Fair Diffusion: Instructing Text-to-Image Generation Models on Fairnesscited by 149
  11. 2019Fairness-Aware Ranking in Search & Recommendation Systems with Application to LinkedIn Talent Searchcited by 137
  12. 2023Social Biases through the Text-to-Image Generation Lenscited by 126
  13. 2023Machine Learning for Synthetic Data Generation: A Reviewcited by 118
  14. 2021Salient ImageNet: How to discover spurious features in Deep Learning?cited by 114
  15. 2021Understanding and Evaluating Racial Biases in Image Captioningcited by 108
  16. 2021Are Gender-Neutral Queries Really Gender-Neutral? Mitigating Gender Bias in Image Searchcited by 90
  17. 2019FairFace: Face Attribute Dataset for Balanced Race, Gender, and Agecited by 87
  18. 2019Toward Gender-Inclusive Coreference Resolutioncited by 84
  19. 2021A Survey on Bias in Visual Datasetscited by 84
  20. 2023ALIP: Adaptive Language-Image Pre-training with Synthetic Captioncited by 70
  21. 2020Robustness to Spurious Correlations in Text Classification via Automatically Generated Counterfactualscited by 67
  22. 2022Contrastive Adapters for Foundation Model Group Robustnesscited by 62
  23. 2022Towards Intersectionality in Machine Learning: Including More Identities, Handling Underrepresentation, and Performing Evaluationcited by 59
  24. 2023Uncurated Image-Text Datasets: Shedding Light on Demographic Biascited by 51
  25. 2023Bias Against 93 Stigmatized Groups in Masked Language Models and Downstream Sentiment Classification Taskscited by 51
  26. 2023T2IAT: Measuring Valence and Stereotypical Biases in Text-to-Image Generationcited by 39
  27. 2023COCO-Counterfactuals: Automatically Constructed Counterfactual Examples for Image-Text Pairscited by 39
  28. 2021Exploring the Efficacy of Automatically Generated Counterfactuals for Sentiment Analysiscited by 30
  29. 2022NeuroCounterfactuals: Beyond Minimal-Edit Counterfactuals for Richer Data Augmentationcited by 23
  30. 2022Flexible text generation for counterfactual fairness probingcited by 13
  31. 2023Auditing and Generating Synthetic Data with Controllable Trust Trade-offscited by 13
  32. 2023Bias in Pruned Vision Models: In-Depth Analysis and Countermeasurescited by 11
  33. 2023Masking Strategies for Background Bias Removal in Computer Vision Modelscited by 9
  34. 2023A Multidimensional Analysis of Social Biases in Vision Transformerscited by 8
  35. 2023Run Like a Girl! Sports-Related Gender Bias in Language and Visioncited by 6

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 13; 2025: 27; 2026: 13.

53 papers.

Sort:

By area:

  1. 2024GenderBias-VL: Benchmarking Gender Bias in Vision Language Models via Counterfactual Probingcited by 31
  2. 2024Uncovering Bias in Large Vision-Language Models at Scale with Counterfactualscited by 27
  3. 2025HumaniBench: A Human-Centric Framework for Large Multimodal Models Evaluationcited by 19
  4. 2025MIRROR: Multimodal Cognitive Reframing Therapy for Rolling with Resistancecited by 13
  5. 2024SK-VQA: Synthetic Knowledge Generation at Scale for Training Context-Augmented Multimodal LLMscited by 12
  6. 2025A Large Scale Analysis of Gender Biases in Text-to-Image Generative Modelscited by 12
  7. 2024Social Perception of Faces in a Vision-Language Modelcited by 10
  8. 2025BBQ-V: Benchmarking Visual Stereotype Bias in Large Multimodal Modelscited by 10
  9. 2024Uncovering Bias in Large Vision-Language Models with Counterfactualscited by 9
  10. 2024Social Debiasing for Fair Multi-modal LLMscited by 8
  11. 2024Natural Language Inference Improves Compositionality in Vision-Language Modelscited by 8
  12. 2025VLDBench Evaluating Multimodal Disinformation with Regulatory Alignmentcited by 8
38 more
  1. 2025Bias Is a Subspace, Not a Coordinate: A Geometric Rethinking of Post-hoc Debiasing in Vision-Language Modelscited by 7
  2. 2025Bias in Gender Bias Benchmarks: How Spurious Features Distort Evaluationcited by 5
  3. 2025Beyond Spurious Signals: Debiasing Multimodal Large Language Models via Counterfactual Inference and Adaptive Expert Routingcited by 5
  4. 2024Lookism: The overlooked bias in computer visioncited by 4
  5. 2025Implicit Bias Injection Attacks against Text-to-Image Diffusion Modelscited by 4
  6. 2025Unveiling the "Fairness Seesaw": Discovering and Mitigating Gender and Race Bias in Vision-Language Modelscited by 4
  7. 2025Investigating VLM Hallucination from a Cognitive Psychology Perspective: A First Step Toward Interpretation with Intriguing Observationscited by 4
  8. 2025Investigating Intersectional Bias in Large Language Models using Confidence Disparities in Coreference Resolutioncited by 3
  9. 2025Bias in the Picture: Benchmarking VLMs with Social-Cue News Images and LLM-as-Judge Assessmentcited by 3
  10. 2025DSO: Direct Steering Optimization for Bias Mitigationcited by 3
  11. 2024CounterFace: A Synthetic Face Dataset for Fine-Grained Counterfactual Evaluation of Face Recognition Systemscited by 2
  12. 2024Is What You Ask For What You Get? Investigating Concept Associations in Text-to-Image Modelscited by 2
  13. 2025BiasConnect: Investigating Bias Interactions in Text-to-Image Modelscited by 2
  14. 2025Mitigate One, Skew Another? Tackling Intersectional Biases in Text-to-Image Modelscited by 2
  15. 2025Cultural Awareness in Vision-Language Models: A Cross-Country Explorationcited by 2
  16. 2025From Individuals to Interactions: Benchmarking Gender Bias in Multimodal Large Language Models from the Lens of Social Relationshipcited by 2
  17. 2025OVFact: Measuring and Improving Open-Vocabulary Factuality for Long Caption Modelscited by 2
  18. 2025Bias Beyond Demographics: Probing Decision Boundaries in Black-Box LVLMs via Counterfactual VQAcited by 2
  19. 2025From Preferences to Prejudice: The Role of Alignment Tuning in Shaping Social Bias in Video Diffusion Modelscited by 2
  20. 2025Representation-Level Counterfactual Calibration for Debiased Zero-Shot Recognitioncited by 2
  21. 2026Measuring Social Bias in Vision-Language Models with Face-Only Counterfactuals from Real Photoscited by 2
  22. 2026Cultural Counterfactuals: Evaluating Cultural Biases in Large Vision-Language Models with Counterfactual Examplescited by 2
  23. 2024Refusal as Silence: Gendered Disparities in Vision-Language Model Responsescited by 1
  24. 2024Evaluation of Attribution Bias in Generator-Aware Retrieval-Augmented Large Language Modelscited by 1
  25. 2025Pruning the Paradox: How CLIP's Most Informative Heads Enhance Performance While Amplifying Biascited by 1
  26. 2025Mapping User Trust in Vision Language Models: Research Landscape, Challenges, and Prospectscited by 1
  27. 2025A Stereotype Content Analysis on Color-related Social Bias in Large Vision Language Modelscited by 1
  28. 2025Surgeons Are Indian Males and Speech Therapists Are White Females: Auditing Biases in Vision-Language Models for Healthcare Professionalscited by 1
  29. 2026Bias at the End of the Scorecited by 1
  30. 2026CounterCount: A Diagnostic Framework for Counting Bias in Vision Language Modelscited by 1
  31. 2026Vision-Language Models Suppress Female Representations Under Ambiguous Inputcited by 1
  32. 2024Debias your Large Multi-Modal Model at Test-Time via Non-Contrastive Visual Attribute Steeringcited by 0
  33. 2025Is GPT-4o mini Blinded by its Own Safety Filters? Exposing the Multimodal-to-Unimodal Bottleneck in Hate Speech Detectioncited by 0
  34. 2026Enhancing Lightweight Vision Language Models through Group Competitive Learning for Socially Compliant Navigationcited by 0
  35. 2026Cross-Cultural Value Attribution in Large Vision-Language Modelscited by 0
  36. 2026Beyond Perception Errors: Semantic Fixation in Large Vision-Language Modelscited by 0
  37. 2026Selective Test-Time Debiasing for CLIP via Reward Gatingcited by 0
  38. 2026Symphony of Bias: Exploring Gender Associations with Musical Instruments in Multimodal LLMscited by 0

50 most cited of 53 All 53 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2023Probing Intersectional Biases in Vision-Language Models with Counterfactual Examples
  2. 2024MIST: Mitigating Intersectional Bias with Disentangled Cross-Attention Editing in Text-to-Image Diffusion Models
  3. 2024BiasDora: Exploring Hidden Biased Associations in Vision-Language Models
  4. 2024Identifying Implicit Social Biases in Vision-Language Models
  5. 2024A Unified Framework and Dataset for Assessing Societal Bias in Vision-Language Models
  6. 2025Data Matters Most: Auditing Social Bias in Contrastive Vision Language Models
  7. 2025Visual Cues of Gender and Race are Associated with Stereotyping in Vision-Language Models
  8. 2023Exploring Social Bias in Downstream Applications of Text-to-Image Foundation Models