arXiv 2603.066402026-02-28cs.CVcs.LG

剪枝之下：揭示基于剪枝的去学习中概念复兴的风险

Roots Beneath the Cut: Uncovering the Risk of Concept Revival in Pruning-Based Unlearning for Diffusion Models

University of Georgia（佐治亚大学）；Carnegie Mellon University（卡内基梅隆大学）；Northeastern University（东北大学）；Stevens Institute of Technology（史蒂文斯理工学院）；University of Arizona（亚利桑那大学）

Ci Zhang, Zhaojun Ding, Chence Yang, Jun Liu, Xiaoming Zhai, Shaoyi Huang, Beiwen Li, Xiaolong Ma, Jin Lu, Geng Yuan

备注：: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2026

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

本文揭示基于剪枝的去学习中概念复兴的风险，提出攻击框架可无数据恢复被擦除概念，并探讨安全剪枝机制。

摘要：

基于剪枝的去学习最近出现了一种快速、无需训练、数据无关的方法，用于从扩散模型中移除不希望的概念。它承诺高效和稳健，提供了一种有吸引力的替代传统微调或编辑-based去学习的方法。然而，在本文中，我们揭示了这种有前途范式背后隐藏的危险。我们发现，在去学习过程中通常被设置为零的剪枝权重的位置可以作为侧通道信号，泄露关于被擦除概念的关键信息。为了验证这种漏洞，我们设计了一个新的攻击框架，能够以完全数据无关和无需训练的方式从剪枝的扩散模型中恢复被擦除的概念。我们的实验确认，基于剪枝的去学习并不本质上安全，因为被擦除的概念可以有效地被恢复，而无需任何额外的数据或重新训练。广泛的实验表明，基于概念相关权重的扩散模型去学习得出的结论是：一旦确定了扩散模型中关键的概念相关权重，我们的方法可以有效地恢复原始概念，无论权重如何被操作。最后，我们探讨了潜在的防御策略，并倡导更安全的剪枝机制，这些机制在隐藏剪枝位置的同时保持去学习的有效性，为设计更安全的基于剪枝的去学习框架提供了实用见解。

英文摘要：

Pruning-based unlearning has recently emerged as a fast, training-free, and data-independent approach to remove undesired concepts from diffusion models. It promises high efficiency and robustness, offering an attractive alternative to traditional fine-tuning or editing-based unlearning. However, in this paper we uncover a hidden danger behind this promising paradigm. We find that the locations of pruned weights, typically set to zero during unlearning, can act as side-channel signals that leak critical information about the erased concepts. To verify this vulnerability, we design a novel attack framework capable of reviving erased concepts from pruned diffusion models in a fully data-free and training-free manner. Our experiments confirm that pruning-based unlearning is not inherently secure, as erased concepts can be effectively revived without any additional data or retraining. Extensive experiments on diffusion-based unlearning based on concept related weights lead to the conclusion: once the critical concept-related weights in diffusion models are identified, our method can effectively recover the original concept regardless of how the weights are manipulated. Finally, we explore potential defense strategies and advocate safer pruning mechanisms that conceal pruning locations while preserving unlearning effectiveness, providing practical insights for designing more secure pruning-based unlearning frameworks.

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。