Vipera: Blending Visual and LLM-Driven Guidance for Systematic Auditing of Text-to-Image Generative AI

要旨

Despite their increasing capabilities, text-to-image generative AI systems are known to produce biased, offensive, and otherwise problematic outputs. While recent advancements have supported testing and auditing of generative AI, existing auditing methods still face challenges in supporting effective exploration of the vast space of AI-generated outputs in a structured way. To address this gap, we conducted formative studies with five AI auditors and synthesized five design goals for supporting systematic AI audits. Based on these insights, we developed Vipera, an interactive auditing interface that employs multiple visual cues, including a scene graph, to facilitate image sensemaking and inspire auditors to explore and hierarchically organize the auditing criteria. Additionally, Vipera leverages LLM-powered suggestions to enable exploration of unexplored auditing directions. Through a controlled experiment with 24 participants experienced in AI auditing, we demonstrate Vipera’s effectiveness in helping auditors navigate large AI output spaces and organize their analyses while engaging with diverse criteria.

著者
Yanwei Huang
HKUST, Hong Kong S.A.R., China
Wesley Hanwen. Deng
Carnegie Mellon University, Pittsburgh, Pennsylvania, United States
Sijia Xiao
Carnegie Mellon University, Pittsburgh, Pennsylvania, United States
Motahhare Eslami
Carnegie Mellon University, Pittsburgh, Pennsylvania, United States
Jason I. Hong
Carnegie Mellon University, Pittsburgh, Pennsylvania, United States
Arpit Narechania
The Hong Kong University of Science and Technology, Clear Water Bay, Hong Kong
Adam Perer
Carnegie Mellon University, Pittsburgh, Pennsylvania, United States

会議: CHI 2026

ACM CHI Conference on Human Factors in Computing Systems

セッション: Generative AI in Design and Practice

P1 - Room 116
7 件の発表
2026-04-17 20:15:00
2026-04-17 21:45:00