Cover Image

Locating Demographic Bias at the Attention-Head Level in CLIP's Vision Encoder

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Yasser, Alaa, Phunjanna, Kittipat, Viñolo, Marcos Escudero, Barata, Catarina, Benois-Pineau, Jenny
Format:	Preprint
Published:	2026
Subjects:	Computer Vision and Pattern Recognition Artificial Intelligence Computers and Society
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

There is More to Attention: Statistical Filtering Enhances Explanations in Vision Transformers
by: Ayyar, Meghna P, et al.
Published: (2025)

Identifying Surgical Instruments in Laparoscopy Using Deep Learning Instance Segmentation
by: Kletz, Sabrina, et al.
Published: (2025)

Pinpoint Counterfactuals: Reducing social bias in foundation models via localized counterfactual generation
by: Sirotkin, Kirill, et al.
Published: (2024)

Demographic Bias of Expert-Level Vision-Language Foundation Models in Medical Imaging
by: Yang, Yuzhe, et al.
Published: (2024)

Object segmentation in the wild with foundation models: application to vision assisted neuro-prostheses for upper limbs
by: Atoki, Bolutife, et al.
Published: (2025)

Debiasing CLIP: Interpreting and Correcting Bias in Attention Heads
by: Yeo, Wei Jie, et al.
Published: (2025)

Metrics for Dataset Demographic Bias: A Case Study on Facial Expression Recognition
by: Dominguez-Catena, Iris, et al.
Published: (2023)

Demographic and Linguistic Bias Evaluation in Omnimodal Language Models
by: Elobaid, Alaa
Published: (2026)

FusWay: Multimodal hybrid fusion approach. Application to Railway Defect Detection
by: Zhukov, Alexey, et al.
Published: (2025)

Mean Opinion Score as a New Metric for User-Evaluation of XAI Methods
by: Yu, Hyeon, et al.
Published: (2024)

SatCLIP: Global, General-Purpose Location Embeddings with Satellite Imagery
by: Klemmer, Konstantin, et al.
Published: (2023)

Vision-Language Models for Autonomous Driving: CLIP-Based Dynamic Scene Understanding
by: Elhenawy, Mohammed, et al.
Published: (2025)

Sum of Group Error Differences: A Critical Examination of Bias Evaluation in Biometric Verification and a Dual-Metric Measure
by: Elobaid, Alaa, et al.
Published: (2024)

A Unified Framework and Dataset for Assessing Societal Bias in Vision-Language Models
by: Sathe, Ashutosh, et al.
Published: (2024)

Impact of Blur and Resolution on Demographic Disparities in 1-to-Many Facial Identification
by: Bhatta, Aman, et al.
Published: (2023)

Generated Bias: Auditing Internal Bias Dynamics of Text-To-Image Generative Models
by: Mandal, Abhishek, et al.
Published: (2024)

Soft labelling for semantic segmentation: Bringing coherence to label down-sampling
by: Alcover-Couso, Roberto, et al.
Published: (2023)

Examining Gender and Racial Bias in Large Vision-Language Models Using a Novel Dataset of Parallel Images
by: Fraser, Kathleen C., et al.
Published: (2024)

Joint Vision-Language Social Bias Removal for CLIP
by: Zhang, Haoyu, et al.
Published: (2024)

Ethical Challenges in Computer Vision: Ensuring Privacy and Mitigating Bias in Publicly Available Datasets
by: Tahir, Ghalib Ahmed
Published: (2024)

Detecting Visual Triggers in Cannabis Imagery: A CLIP-Based Multi-Labeling Framework with Local-Global Aggregation
by: Lu, Linqi, et al.
Published: (2024)

Gradient-based Class Weighting for Unsupervised Domain Adaptation in Dense Prediction Visual Tasks
by: Alcover-Couso, Roberto, et al.
Published: (2024)

VLMs meet UDA: Boosting Transferability of Open Vocabulary Segmentation with Unsupervised Domain Adaptation
by: Alcover-Couso, Roberto, et al.
Published: (2024)

Attention Head Purification: A New Perspective to Harness CLIP for Domain Generalization
by: Wang, Yingfan, et al.
Published: (2024)

MIL vs. Aggregation: Evaluating Patient-Level Survival Prediction Strategies Using Graph-Based Learning
by: Verdelho, M Rita, et al.
Published: (2025)

Beyond the Vision Encoder: Identifying and Mitigating Spatial Bias in Large Vision-Language Models
by: Zhu, Yingjie, et al.
Published: (2025)

Leveraging CLIP Encoder for Multimodal Emotion Recognition
by: Song, Yehun, et al.
Published: (2025)

Open-Vocabulary X-ray Prohibited Item Detection via Fine-tuning CLIP
by: Lin, Shuyang, et al.
Published: (2024)

Attention is All They Need: Exploring the Media Archaeology of the Computer Vision Research Paper
by: Goree, Samuel, et al.
Published: (2022)

FairREAD: Re-fusing Demographic Attributes after Disentanglement for Fair Medical Image Classification
by: Gao, Yicheng, et al.
Published: (2024)

KG-FairDiff: Knowledge Graph-Guided Prompt Refinement for Demographically Fair Text-to-Image Generation
by: Davoodi, Farbod, et al.
Published: (2026)

Happy Young Women, Grumpy Old Men? Emotion-Driven Demographic Biases in Synthetic Face Generation
by: Wei, Mengting, et al.
Published: (2026)

Improving Vision Transformers by Overlapping Heads in Multi-Head Self-Attention
by: Zhang, Tianxiao, et al.
Published: (2024)

Attention to Neural Plagiarism: Diffusion Models Can Plagiarize Your Copyrighted Images!
by: Zou, Zihang, et al.
Published: (2026)

Multimodal Political Bias Identification and Neutralization
by: Bernard, Cedric, et al.
Published: (2025)

Exploring How Generative MLLMs Perceive More Than CLIP with the Same Vision Encoder
by: Li, Siting, et al.
Published: (2024)

Certified Circuits: Stability Guarantees for Mechanistic Circuits
by: Anani, Alaa, et al.
Published: (2026)

Leveraging Contrastive Learning for Semantic Segmentation with Consistent Labels Across Varying Appearances
by: Montalvo, Javier, et al.
Published: (2024)

Can CLIP Count Stars? An Empirical Study on Quantity Bias in CLIP
by: Zhang, Zeliang, et al.
Published: (2024)

Is CLIP Cross-Eyed? Revealing and Mitigating Center Bias in the CLIP Family
by: Chew, Oscar, et al.
Published: (2026)