不稳定的遗忘：扩散模型中概念复苏的隐藏风险
Unstable Unlearning: The Hidden Risk of Concept Resurgence in Diffusion Models
AI总结：
研究发现，对文本到图像扩散模型进行良性微调可导致已遗忘概念复苏，揭示增量更新组合的脆弱性，对模型安全对齐提出新挑战。
摘要：
文本到图像扩散模型依赖于大规模的网络级数据集。从头训练这些模型在计算上代价高昂，因此开发者通常倾向于对现有模型进行增量更新。这些更新常常将微调步骤（用于学习新概念或提升模型性能）与“遗忘”步骤（用于“忘记”现有概念，如受版权保护的作品或露骨内容）组合在一起。在这项工作中，我们展示了一个在此范式中出现的关键且此前未知的漏洞：即使在良性、非对抗性条件下，对文本到图像扩散模型在看似无关的图像上进行微调，也可能导致其“重新学习”先前已被“遗忘”的概念。我们通过一系列实验，将“概念遗忘”与随后对Stable Diffusion v1.4和Stable Diffusion v2.1的微调相结合，全面调查了这一我们称之为“概念复苏”现象的原因和范围。我们的发现强调了组合增量模型更新的脆弱性，并对当前确保文本到图像扩散模型安全性和对齐的方法提出了严重的新担忧。
英文摘要：
Text-to-image diffusion models rely on massive, web-scale datasets. Training them from scratch is computationally expensive, and as a result, developers often prefer to make incremental updates to existing models. These updates often compose fine-tuning steps (to learn new concepts or improve model performance) with "unlearning" steps (to "forget" existing concepts, such as copyrighted works or explicit content). In this work, we demonstrate a critical and previously unknown vulnerability that arises in this paradigm: even under benign, non-adversarial conditions, fine-tuning a text-to-image diffusion model on seemingly unrelated images can cause it to "relearn" concepts that were previously "unlearned." We comprehensively investigate the causes and scope of this phenomenon, which we term concept resurgence, by performing a series of experiments which compose "concept unlearning" with subsequent fine-tuning of Stable Diffusion v1.4 and Stable Diffusion v2.1. Our findings underscore the fragility of composing incremental model updates, and raise serious new concerns about current approaches to ensuring the safety and alignment of text-to-image diffusion models.
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。
