aiqbench analysis
BioPro introduces a training-free framework for vision-language models that selectively neutralizes gender bias in irrelevant contexts while preserving legitimate distinctions where necessary. This approach addresses the critical nuance between unwanted stereotypes and valid demographic attributes, offering a practical path toward more equitable multimodal AI systems without requiring model retraining. The method also demonstrates generalizability to other continuous variables like scene brightness, suggesting broader utility for fairness interventions across different bias types.
BioPro: Towards Difference-Aware Gender Fairness for Vision-Language Models
arXiv:2512.00807v2 Announce Type: replace Abstract: Vision-Language Models (VLMs) inherit significant social biases from their training data, notably in gender representation. Current fairness interventions often adopt a difference-unaware perspective that enforces uniform treatment across demographic groups. These approaches, however, fail to distinguish between contexts where neutrality is required and those where group-specific attributes are legitimate and must be preserved. Building upon recent advances in difference-aware fairness for text-only models, we extend this concept to the multimodal domain and formalize the problem of difference-aware gender fairness for image captioning and text-to-image generation. We advocate for selective debiasing, which aims to mitigate unwanted bias in neutral contexts while preserving valid distinctions in explicit ones. To achieve this, we propose BioPro (Bias Orthogonal Projection), an entirely training-free framework. BioPro identifies a low-dimensional gender-variation subspace through counterfactual embeddings and applies projection to selectively neutralize gender-related information. Experiments show that BioPro effectively reduces gender bias in neutral cases while maintaining gender faithfulness in explicit ones, thus providing a promising direction toward achieving selective fairness in VLMs. Beyond gender bias, we further demonstrate that BioPro can effectively generalize to continuous bias variables, such as scene brightness, highlighting its broader applicability.
Abstract:Vision-Language Models (VLMs) inherit significant social biases from their training data, notably in gender representation. Current fairness interventions often adopt a difference-unaware perspective that enforces uniform treatment across demographic groups. These approaches, however, fail to distinguish between contexts where neutrality is required and those where group-specific attributes are legitimate and must be preserved. Building upon recent advances in difference-aware fairness for text-only models, we extend this concept to the multimodal domain and formalize the problem of difference-aware gender fairness for image captioning and text-to-image generation. We advocate for selective debiasing, which aims to mitigate unwanted bias in neutral contexts while preserving valid distinctions in explicit ones. To achieve this, we propose BioPro (Bias Orthogonal Projection), an entirely training-free framework. BioPro identifies a low-dimensional gender-variation subspace through counterfactual embeddings and applies projection to selectively neutralize gender-related information. Experiments show that BioPro effectively reduces gender bias in neutral cases while maintaining gender faithfulness in explicit ones, thus providing a promising direction toward achieving selective fairness in VLMs. Beyond gender bias, we further demonstrate that BioPro can effectively generalize to continuous bias variables, such as scene brightness, highlighting its broader applicability.
Artificial Intelligence (cs.AI)
arXiv:2512.00807 [cs.AI]
(or arXiv:2512.00807v2 [cs.AI] for this version)
https://doi.org/10.48550/arXiv.2512.00807 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 280 words
Related coverage
AI Futures: Racing to Lose, A sermon
LessWrong · Sep 28, 2026
BRICS Draws a New Global Line: Full Text of New Delhi Declaration
slguardian.org · Sep 12, 2026
Digital States 2026: AI Moves From Experiment to Infrastructure
govtech.com · Sep 24, 2026
The world's best gradual disempowerment model organism: Frontier AI labs
LessWrong · Sep 29, 2026