利用反事实大规模揭示大型视觉语言模型中的偏差
Uncovering Bias in Large Vision-Language Models at Scale with Counterfactuals
AI总结：
本文通过反事实改变输入图像，对多个大型视觉语言模型（LVLMs）生成的大规模文本进行研究，提出多维度偏差评估框架，揭示图像中的种族、性别等社会属性对生成内容偏差的显著影响。
摘要：
随着具备日益惊人能力的大型语言模型（LLMs）的出现，许多大型视觉语言模型（LVLMs）被提出以通过视觉输入增强 LLMs。此类模型基于输入图像和文本提示来生成文本，实现了视觉问答和多模态聊天等多种应用。尽管先前的研究已检验了 LLMs 生成文本中包含的社会偏差，但这一主题在 LVLMs 中相对未被探索。由于文本和视觉模态中所含信息引发的偏差存在混淆作用，检验 LVLMs 中的社会偏差尤为困难。为解决这一难题，我们在对输入图像进行反事实改变的条件下，对不同 LVLMs 生成的文本进行了大规模研究，从流行模型中产生了超过 5700 万条响应。我们的多维度偏差评估框架揭示，图像中描绘的感知种族、性别和身体特征等社会属性，能够显著影响有毒内容、能力关联词、有害刻板印象的生成，以及对个体的数值评分。
英文摘要：
With the advent of Large Language Models (LLMs) possessing increasingly impressive capabilities, a number of Large Vision-Language Models (LVLMs) have been proposed to augment LLMs with visual inputs. Such models condition generated text on both an input image and a text prompt, enabling a variety of use cases such as visual question answering and multimodal chat. While prior studies have examined the social biases contained in text generated by LLMs, this topic has been relatively unexplored in LVLMs. Examining social biases in LVLMs is particularly challenging due to the confounding contributions of bias induced by information contained across the text and visual modalities. To address this challenging problem, we conduct a large-scale study of text generated by different LVLMs under counterfactual changes to input images, producing over 57 million responses from popular models. Our multi-dimensional bias evaluation framework reveals that social attributes such as perceived race, gender, and physical characteristics depicted in images can significantly influence the generation of toxic content, competency-associated words, harmful stereotypes, and numerical ratings of individuals.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
