Embedding Arithmetic performs vector operations in the embedding space of T2I models to mitigate bias at inference time, outperforming baselines on diversity while preserving coherence via a new Concept Coherence Score.
Debias- ing vision-language models via biased prompts.arXiv preprint arXiv:2302.00070
4 Pith papers cite this work. Polarity classification is still indexing.
4
Pith papers citing it
verdicts
UNVERDICTED 4representative citing papers
POVID generates AI-created preference data to fine-tune vision-language models with DPO, reducing hallucinations and improving benchmark scores.
Vision-language models for wellbeing assessment exhibit dataset-dependent performance and demographic biases, with explainability interventions providing inconsistent fairness gains at potential accuracy costs.
A systematic review of T2I bias literature that distinguishes target and threshold fairness and proposes a target-based operationalization framework.
citing papers explorer
-
Embedding Arithmetic: A Lightweight, Tuning-Free Framework for Post-hoc Bias Mitigation in Text-to-Image Models
Embedding Arithmetic performs vector operations in the embedding space of T2I models to mitigate bias at inference time, outperforming baselines on diversity while preserving coherence via a new Concept Coherence Score.
-
Aligning Modalities in Vision Large Language Models via Preference Fine-tuning
POVID generates AI-created preference data to fine-tune vision-language models with DPO, reducing hallucinations and improving benchmark scores.
-
FAIR_XAI: Improving Multimodal Foundation Model Fairness via Explainability for Wellbeing Assessment
Vision-language models for wellbeing assessment exhibit dataset-dependent performance and demographic biases, with explainability interventions providing inconsistent fairness gains at potential accuracy costs.
-
Operationalizing Fairness in Text-to-Image Models: A Survey of Bias, Fairness Audits and Mitigation Strategies
A systematic review of T2I bias literature that distinguishes target and threshold fairness and proposes a target-based operationalization framework.