arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2308.09971cs.LGcs.AIcs.CV
收藏

面向选择性源任务遗忘的一次性迁移学习

Disposable Transfer Learning for Selective Source Task Unlearning

Korea Advanced Institute of Science and Technology(韩国科学技术院);LG AI Research(LG人工智能研究院)

Seunghee Koh, Hyounguk Shon, Janghyeon Lee, Hyeong Gwon Hong, Junmo Kim

AI总结:

为保护预训练模型性能的私有性,提出一次性迁移学习范式与梯度碰撞损失,可选择性遗忘源任务知识且不降低目标任务性能,经搭载学习准确率验证效果显著。

摘要:

迁移学习被广泛用于训练深度神经网络(DNN)以构建强大的表征。即便预训练模型已适配目标任务,特征提取器的表征性能仍会在一定程度上被保留。由于预训练模型的性能可被视为其所有者的私有财产,寻求预训练权重泛化性能的专有权是很自然的需求。为解决这一问题,我们提出了一种名为一次性迁移学习(disposable transfer learning,DTL)的迁移学习新范式,该范式仅清除源任务相关知识,同时不会降低目标任务的性能。为实现知识清除,我们提出了一种名为梯度碰撞损失(Gradient Collision loss,GC loss)的新型损失函数。GC loss通过引导小批量样本的梯度向量朝向不同方向,实现源知识的选择性遗忘。模型是否成功遗忘源任务,通过搭载学习准确率(piggyback learning accuracy,PL accuracy)来衡量。PL accuracy通过在源数据子集或新的下游数据上重新训练清洗后的模型,来评估知识泄露的脆弱性。我们通过实验证明,采用GC loss训练的模型在保留目标任务性能的同时,PL accuracy显著降低,验证了GC loss是解决DTL问题的有效方法。

英文摘要:

Transfer learning is widely used for training deep neural networks (DNN) for building a powerful representation. Even after the pre-trained model is adapted for the target task, the representation performance of the feature extractor is retained to some extent. As the performance of the pre-trained model can be considered the private property of the owner, it is natural to seek the exclusive right of the generalized performance of the pre-trained weight. To address this issue, we suggest a new paradigm of transfer learning called disposable transfer learning (DTL), which disposes of only the source task without degrading the performance of the target task. To achieve knowledge disposal, we propose a novel loss named Gradient Collision loss (GC loss). GC loss selectively unlearns the source knowledge by leading the gradient vectors of mini-batches in different directions. Whether the model successfully unlearns the source task is measured by piggyback learning accuracy (PL accuracy). PL accuracy estimates the vulnerability of knowledge leakage by retraining the scrubbed model on a subset of source data or new downstream data. We demonstrate that GC loss is an effective approach to the DTL problem by showing that the model trained with GC loss retains the performance on the target task with a significantly reduced PL accuracy.

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。

↑