论深度持续学习中灾难性遗忘的隐式对抗性
On the Implicit Adversariality of Catastrophic Forgetting in Deep Continual Learning
AI总结：
本文揭示持续学习中灾难性遗忘源于新任务梯度对旧任务损失景观尖锐方向的隐式对抗性对齐，归因于训练的低秩偏差，并提出backGP方法缓解该问题，平均减少遗忘10.8%、提升准确率12.7%。
摘要：
持续学习旨在使机器智能具备类似人类积累新技能的能力。其核心挑战是灾难性遗忘，而对于深度网络而言，其根本原因尚未被完全理解。本文通过揭示新任务训练本质上是对旧任务知识的隐式对抗攻击，来阐明灾难性遗忘的机理。具体而言，新任务的梯度会自动且准确地与旧任务损失景观的尖锐方向对齐，从而迅速增加旧任务的损失。这种对抗性对齐令人惊讶地反直觉，因为尖锐方向分布极为稀疏，偶然对齐的可能性极低。为理解这一现象，我们从理论上证明其源于训练的低秩偏差，该偏差通过前向和后向传播将两个方向限制在同一低维子空间中，从而促进对齐。梯度投影（GP）方法作为缓解遗忘方法的代表性家族，能减少由前向传播引起的对抗性对齐，但无法解决由后向传播导致的对齐问题。我们提出backGP来解决这一问题，与GP方法相比，平均减少遗忘10.8%，并提高准确率12.7%。
英文摘要：
Continual learning seeks the human-like ability to accumulate new skills in machine intelligence. Its central challenge is catastrophic forgetting, whose underlying cause has not been fully understood for deep networks. In this paper, we demystify catastrophic forgetting by revealing that the new-task training is implicitly an adversarial attack against the old-task knowledge. Specifically, the new-task gradients automatically and accurately align with the sharp directions of the old-task loss landscape, rapidly increasing the old-task loss. This adversarial alignment is intriguingly counter-intuitive because the sharp directions are too sparsely distributed to align with by chance. To understand it, we theoretically show that it arises from training's low-rank bias, which, through forward and backward propagation, confines the two directions into the same low-dimensional subspace, facilitating alignment. Gradient projection (GP) methods, a representative family of forgetting-mitigating methods, reduce adversarial alignment caused by forward propagation, but cannot address the alignment due to backward propagation. We propose backGP to address it, which reduces forgetting by 10.8% and improves accuracy by 12.7% on average over GP methods.
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。
