视觉-语言模型中的Déjà Vu记忆现象
Déjà Vu Memorization in Vision-Language Models
AI总结:
该研究针对视觉-语言模型是否记忆训练数据的问题,提出déjà vu记忆衡量方法,证实OpenCLIP存在显著记忆效应,并发现文本随机化可有效缓解该问题且对下游性能影响较小。
摘要:
视觉-语言模型(VLMs)已成为当前最先进的表征学习解决方案,拥有图像分类、检索和生成等大量下游应用。一个自然的问题是这些模型是否会记忆其训练数据,这对模型的泛化能力也有影响。我们提出了一种衡量VLM中记忆程度的新方法,将其称为déjà vu记忆。针对在图像-字幕对上训练的VLM,我们证明模型确实保留了训练图像中单个物体的信息,且这些信息超出了可从相关性或图像字幕推断出的范围。我们在样本层面和群体层面对déjà vu记忆进行了评估,结果表明,对于在多达5000万对图像-字幕对上训练的OpenCLIP,这种记忆效应十分显著。最后,我们发现文本随机化能大幅缓解记忆问题,同时仅会对模型的下游任务性能产生中等程度的影响。
英文摘要:
Vision-Language Models (VLMs) have emerged as the state-of-the-art representation learning solution, with myriads of downstream applications such as image classification, retrieval and generation. A natural question is whether these models memorize their training data, which also has implications for generalization. We propose a new method for measuring memorization in VLMs, which we call déjà vu memorization. For VLMs trained on image-caption pairs, we show that the model indeed retains information about individual objects in the training images beyond what can be inferred from correlations or the image caption. We evaluate déjà vu memorization at both sample and population level, and show that it is significant for OpenCLIP trained on as many as 50M image-caption pairs. Finally, we show that text randomization considerably mitigates memorization while only moderately impacting the model's downstream task performance.
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。