DECAF：用于自适应表征遗忘的去聚类方法
DECAF: De-Clustering for Adaptive Representational Unlearning
浏览论文内容
中文总结 AI 辅助
研究针对机器遗忘中聚类攻击问题，提出DECAF方法，结合输入噪声、置信度抑制和输出多样化破坏遗忘数据特征结构，在CIFAR-10与ResNet-18上表现出色，性能超基线且效率高。
中文摘要 AI 辅助
机器遗忘旨在从训练模型中消除特定训练数据的影响，是隐私、问责制和自适应部署的关键要求。许多遗忘方法易受简单聚类攻击，限制了其在持续部署中的适用性。为此，我们提出了DECAF，一种仅对遗忘集进行操作并旨在打破聚类的事后方法。它结合输入噪声、置信度抑制和基于熵的输出多样化来破坏与遗忘数据相关的残余特征空间结构。在CIFAR-10数据集和ResNet-18模型上，DECAF实现了0.10%的遗忘类准确率、79.4%的保留准确率和0.88的AUS，超过所有基线。在基于聚类的分析中，它达到了与使用完整训练集进行遗忘方法相当的性能，同时效率显著更高。代码：此https链接。
英文摘要
Machine unlearning, which aims to remove the influence of specific training data from a trained model, is a key requirement for privacy, accountability, and adaptive deployment. We argue that many unlearning methods are vulnerable to a simple clustering attack, which can recover class structure in an unsupervised manner, limiting their suitability for continual deployment where removal requests must be handled reliably on demand. To address this, we propose DECAF (DE-Clustering for Adaptive Forgetting), a post-hoc method that operates only on the forget set and is designed to break the cluster. DECAF combines input noise, confidence suppression, and entropy-based output diversification to disrupt the residual feature-space structure associated with forgotten data. On CIFAR-10 with ResNet-18, DECAF attains 0.10% forget-class accuracy, 79.4% retain accuracy, and an AUS of 0.88, surpassing all other baselines. In cluster-based analysis, it attains performance comparable to that of unlearning methods that use the full training set, while being significantly more efficient. Code: https://github.com/ale256/representation_unlearning.
