研究发现主流文生图模型在潜空间中系统性编码了种族与性别偏见。
T2IBias: Uncovering Societal Bias Encoded in the Latent Space of Text-to-Image Generative Models
- 通过十类职业提示生成5000张图像，评估潜空间中的社会偏见
- 所有模型均呈现女性化护理角色、男性白人主导高阶职业的显著偏差
- 揭示不同模型特异性偏见，如通义图像专注东亚面孔
文生图（T2I）生成模型广泛应用于人工智能驱动的实际场景与价值创造。然而其战略部署引发负责任AI管理的重大关切，尤其体现在对种族与性别刻板印象的再现与放大，可能损害组织伦理。本文探究主流T2I模型预训练潜空间是否系统性编码此类社会偏见。我们在五个最流行的开源模型上开展实证研究，使用十组中性职业相关提示，每职业生成100张图像，共构建5000张图像数据集，并由多元种族与性别的真人评估者进行评价。结果表明，所有五款模型均编码并放大显著的社会偏差：照护与护士角色持续女性化，而企业首席执行官、政治家、医生、律师等高地位职业则主要由男性和白人代表。进一步发现模型特异性模式，如QWEN-Image几乎仅输出东亚面孔，Kandinsky以白人为主导，SDXL虽分布较广但仍具偏差。这些发现为AI项目管理者与实践者提供关键洞见，助力选择更公平的AI模型与定制提示，确保生成内容符合负责任AI原则。最后讨论偏见风险并提出可操作的缓解策略。代码与数据仓库：https://github.com/Sufianlab/T2IBias
原文摘要 · Abstract (English)
Text-to-image (T2I) generative models are largely used in AI-powered real-world applications and value creation. However, their strategic deployment raises critical concerns for responsible AI management, particularly regarding the reproduction and amplification of race- and gender-related stereotypes that can undermine organizational ethics. In this work, we investigate whether such societal biases are systematically encoded within the pretrained latent spaces of state-of-the-art T2I models. We conduct an empirical study across the five most popular open-source models, using ten neutral, profession-related prompts to generate 100 images per profession, resulting in a dataset of 5,000 images evaluated by diverse human assessors representing different races and genders. We demonstrate that all five models encode and amplify pronounced societal skew: caregiving and nursing roles are consistently feminized, while high-status professions such as corporate CEO, politician, doctor, and lawyer are overwhelmingly represented by males and mostly White individuals. We further identify model-specific patterns, such as QWEN-Image's near-exclusive focus on East Asian outputs, Kandinsky's dominance of White individuals, and SDXL's comparatively broader but still biased distributions. These results provide critical insights for AI project managers and practitioners, enabling them to select equitable AI models and customized prompts that generate images in alignment with the principles of responsible AI. We conclude by discussing the risks of these biases and proposing actionable strategies for bias mitigation in building responsible GenAI systems. The code and Data Repository: https://github.com/Sufianlab/T2IBias
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。