Interference-Aware Multi-Task Unlearning

论文概要

研究领域: cs.AI 作者: Ying-Hua Huang, Rui Fang, Hsi-Wen Chen 发布时间: 2026-05-21 arXiv: 2505.01257

中文摘要

机器遗忘旨在从训练好的模型中移除指定训练数据的贡献，同时保留对剩余数据的性能。现有工作主要关注单任务设置，而现代模型通常在具有共享骨干的多任务设置中运行，其中移除一个任务或实例的监督可能无意影响其他任务。我们引入了多任务遗忘，包含两种设置：全任务遗忘，从所有任务中移除目标实例；部分任务遗忘，仅从选定任务中移除监督。我们展示了共享参数耦合了遗忘集和保留集，导致非目标任务上的任务级干扰和其他实例上的实例级干扰。为解决此问题，我们提出了一种干扰感知框架，结合任务感知梯度投影（将更新约束在任务特定子空间内）和实例级梯度正交化（减少遗忘和保留信号之间的冲突）。在两个多任务计算机视觉基准测试中，跨五个任务的实验表明，我们的方法在实现有效遗忘的同时保持强泛化能力，与最强基线相比，在全任务遗忘中将UIS降低30.3%，在部分任务遗忘中降低52.9%。

原文摘要

Machine unlearning aims to remove the contribution of designated training data from a trained model while preserving performance on the remaining data. Existing work mainly focuses on single-task settings, whereas modern models often operate in multi-task setups with shared backbones, where removing supervision for one task or instance can unintentionally affect others. We introduce multi-task unlearning with two settings: full-task unlearning, which removes a target instance from all tasks, and partial-task unlearning, which removes supervision only from selected tasks. We show that shared parameters couple the forget and retain sets, causing task-level interference on non-target tasks and instance-level interference on other instances. To address this issue, we propose an interference-aware framework that combines task-aware gradient projection, which constrains updates within task-specific subspaces, with instance-level gradient orthogonalization, which reduces conflicts between forget and retain signals. Experiments on two multi-task computer vision benchmarks across five tasks show that our method achieves effective unlearning while maintaining strong generalization, reducing UIS compared with the strongest baseline by 30.3% in full-task unlearning and 52.9% in partial-task unlearning.

自动采集于 2026-05-21

#论文 #arXiv #AI #小凯