Apr 24
Who Defines Fairness? Target-Based Prompting for Demographic Representation in Generative Models
★★★★★
significance 3/5
Researchers propose a new lightweight framework to mitigate demographic bias in text-to-image models during inference. The method allows users to select specific fairness definitions to adjust the demographic representation of generated images without retraining the model.
Why it matters
Dynamic, inference-time control over demographic representation offers a scalable alternative to the costly retraining required to mitigate systemic bias in generative models.
Tags
#generative ai #bias mitigation #text-to-image #fairness #llmRelated coverage
- Global South OpportunitiesPivotal Research Fellowship 2026 (Q3): AI Safety Research Opportunity - Global South Opportunities
- arXiv cs.AIAn Intelligent Fault Diagnosis Method for General Aviation Aircraft Based on Multi-Fidelity Digital Twin and FMEA Knowledge Enhancement
- arXiv cs.AIPExA: Parallel Exploration Agent for Complex Text-to-SQL
- arXiv cs.AIThe Power of Power Law: Asymmetry Enables Compositional Reasoning
- arXiv cs.AIOn the Existence of an Inverse Solution for Preference-Based Reductions in Argumentation