# Seeing Twice: How Side-by-Side T2I Comparison Changes Auditing Strategies

arXiv: https://arxiv.org/abs/2511.21547 · HTML: https://mlsanity.com/paper/2511.21547/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Vision-language and text-to-image safety, one of 12 directions in LLM safety and security.
- Builds on 6 of its 15 references that are in the map.
- First posted on arXiv 26 Nov 2025.

## Placed in

- LLM safety and security / [Vision-language and text-to-image safety](https://mlsanity.com/safety/vision-language-and-text-to-image-safety/)

## Builds on (most cited first, 6 of 6)

- [Chatbot Arena: An Open Platform for Evaluating LLMs by Human Preference](https://mlsanity.com/paper/2403.04132/) (2024)
- [Everyday algorithm auditing: Understanding the power of everyday users in surfacing harmful algorithmic behaviors](https://mlsanity.com/paper/2105.02980/) (2021)
- [AI auditing: The Broken Bus on the Road to AI Accountability](https://mlsanity.com/paper/2401.14462/) (2024)
- [Understanding Practices, Challenges, and Opportunities for User-Engaged Algorithm Auditing in Industry Practice](https://mlsanity.com/paper/2210.03709/) (2022)
- [WeAudit: Scaffolding User Auditors and AI Practitioners in Auditing Generative AI](https://mlsanity.com/paper/2501.01397/) (2025)
- [MIRAGE: Multi-model Interface for Reviewing and Auditing Generative Text-to-Image AI](https://mlsanity.com/paper/2503.19252/) (2025)

## Abstract (arXiv)

While generative AI systems have gained popularity in diverse applications, their potential to produce harmful outputs limits their trustworthiness and utility. A small but growing line of research has explored tools and processes to better engage non-AI expert users in auditing generative AI systems. In this work, we present the design and evaluation of MIRAGE, a web-based tool exploring a "contrast-first" workflow that allows users to pick up to four different text-to-image (T2I) models, view their images side-by-side, and provide feedback on model performance on a single screen. In our user study with fifteen participants, we used four predefined models for consistency, with only a single model initially being shown. We found that most participants shifted from analyzing individual images to general model output patterns once the side-by-side step appeared with all four models; several participants coined persistent "model personalities" (e.g., cartoonish, saturated) that helped them form expectations about how each model would behave on future prompts. Bilingual participants also surfaced a language-fidelity gap, as English prompts produced more accurate images than Portuguese or Chinese, an issue often overlooked when dealing with a single model. These findings suggest that simple comparative interfaces can accelerate bias discovery and reshape how people think about generative models.