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

TL;DR

GLEaN is a scalable, model-agnostic method that visually explains biases in text-to-image models to the public, using composite portraits that are easy to interpret and compare.

Contribution

Introduces GLEaN, a novel portrait-based explainability pipeline that makes T2I model biases visually understandable for broad audiences, without requiring model internals.

Findings

GLEaN reproduces documented biases in models.

User study shows GLEaN portraits are as effective as data tables in communicating biases.

GLEaN requires less viewing time than traditional bias representations.

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.