Vipera: Towards systematic auditing of generative text-to-image models at scale

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Huang, Yanwei, Deng, Wesley Hanwen, Xiao, Sijia, Eslami, Motahhare, Hong, Jason I., Perer, Adam
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866915197725179904
author Huang, Yanwei
Deng, Wesley Hanwen
Xiao, Sijia
Eslami, Motahhare
Hong, Jason I.
Perer, Adam
author_facet Huang, Yanwei
Deng, Wesley Hanwen
Xiao, Sijia
Eslami, Motahhare
Hong, Jason I.
Perer, Adam
contents Generative text-to-image (T2I) models are known for their risks related such as bias, offense, and misinformation. Current AI auditing methods face challenges in scalability and thoroughness, and it is even more challenging to enable auditors to explore the auditing space in a structural and effective way. Vipera employs multiple visual cues including a scene graph to facilitate image collection sensemaking and inspire auditors to explore and hierarchically organize the auditing criteria. Additionally, it leverages LLM-powered suggestions to facilitate exploration of unexplored auditing directions. An observational user study demonstrates Vipera's effectiveness in helping auditors organize their analyses while engaging with diverse criteria.
format Preprint
id arxiv_https___arxiv_org_abs_2503_11113
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Vipera: Towards systematic auditing of generative text-to-image models at scale
Huang, Yanwei
Deng, Wesley Hanwen
Xiao, Sijia
Eslami, Motahhare
Hong, Jason I.
Perer, Adam
Human-Computer Interaction
Generative text-to-image (T2I) models are known for their risks related such as bias, offense, and misinformation. Current AI auditing methods face challenges in scalability and thoroughness, and it is even more challenging to enable auditors to explore the auditing space in a structural and effective way. Vipera employs multiple visual cues including a scene graph to facilitate image collection sensemaking and inspire auditors to explore and hierarchically organize the auditing criteria. Additionally, it leverages LLM-powered suggestions to facilitate exploration of unexplored auditing directions. An observational user study demonstrates Vipera's effectiveness in helping auditors organize their analyses while engaging with diverse criteria.
title Vipera: Towards systematic auditing of generative text-to-image models at scale
topic Human-Computer Interaction
url https://arxiv.org/abs/2503.11113