GLEaN: A Text-to-image Bias Detection Approach for Public Comprehension

arXiv cs.AI / 4/14/2026

📰 NewsSignals & Early TrendsIdeas & Deep AnalysisTools & Practical UsageModels & Research

共有:

Key Points

GLEaN is a portrait-based, model-agnostic text-to-image bias explainability pipeline aimed at making T2I biases understandable to non-technical audiences.
The method generates images from identity prompts, filters and aligns them using facial landmarks, and then creates a median-pixel composite that visually summarizes the model’s central tendency.
Applied to Stable Diffusion XL across 40 identity prompts, GLEaN reproduces known biases and also surfaces new associations, such as links between skin tone and predicted emotion.
In a user study (N=291), GLEaN portraits communicated bias findings as effectively as conventional tables while significantly reducing the time required to view and interpret results.
Because it uses only generated outputs, GLEaN can be replicated on black-box systems without access to internal model details, and the code is released on GitHub.

Abstract

Text-to-image (T2I) models, and their encoded biases, increasingly shape the visual media the public encounters. While researchers have produced a rich body of work on bias measurement, auditing, and mitigation in T2I systems, those methods largely target technical stakeholders, leaving a gap in public legibility. We introduce GLEaN (Generative Likeness Evaluation at N-Scale), a portrait-based explainability pipeline designed to make T2I model biases visually understandable to a broad audience. GLEaN comprises three stages: automated large-scale image generation from identity prompts, facial landmark-based filtering and spatial alignment, and median-pixel composition that distills a model's central tendency into a single representative portrait. The resulting composites require no statistical background to interpret; a viewer can see, at a glance, who a model 'imagines' when prompted with 'a doctor' versus a 'felon.' We demonstrate GLEaN on Stable Diffusion XL across 40 social and occupational identity prompts, producing composites that reproduce documented biases and surface new associations between skin tone and predicted emotion. We find in a between-subjects user study (N = 291) that GLEaN portraits communicate biases as effectively as conventional data tables, but require significantly less viewing time. Because the method relies solely on generated outputs, it can also be replicated on any black-box and closed-weight systems without access to model internals. GLEaN offers a scalable, model-agnostic approach to bias explainability, purpose-built for public comprehension, and is publicly available at https://github.com/cultureiolab/GLEaN.