Whose View of Safety? A Deep DIVE Dataset for Pluralistic Alignment of Text-to-Image Models
This work addresses the need for more equitable and aligned text-to-image systems by providing foundational tools for pluralistic alignment, though it is incremental in building on existing alignment efforts.
The paper tackles the problem of text-to-image models failing to account for diverse human experiences by introducing the DIVE dataset for pluralistic alignment, which uses demographically intersectional human raters across 1000 prompts to capture nuanced safety perceptions and reveals significant differences in harm perception.
Current text-to-image (T2I) models often fail to account for diverse human experiences, leading to misaligned systems. We advocate for pluralistic alignment, where an AI understands and is steerable towards diverse, and often conflicting, human values. Our work provides three core contributions to achieve this in T2I models. First, we introduce a novel dataset for Diverse Intersectional Visual Evaluation (DIVE) -- the first multimodal dataset for pluralistic alignment. It enable deep alignment to diverse safety perspectives through a large pool of demographically intersectional human raters who provided extensive feedback across 1000 prompts, with high replication, capturing nuanced safety perceptions. Second, we empirically confirm demographics as a crucial proxy for diverse viewpoints in this domain, revealing significant, context-dependent differences in harm perception that diverge from conventional evaluations. Finally, we discuss implications for building aligned T2I models, including efficient data collection strategies, LLM judgment capabilities, and model steerability towards diverse perspectives. This research offers foundational tools for more equitable and aligned T2I systems. Content Warning: The paper includes sensitive content that may be harmful.