用控制理论缓解神经网络持续学习中的遗忘问题
Continual Learning through Control Minimization
- 将持续学习建模为控制问题，通过最小化控制努力实现新旧任务平衡
- 无需存储历史数据或显式计算曲率，即可隐式保留先验任务的曲率信息
- 在不使用回放的情况下超越现有方法，在标准基准上表现优异
灾难性遗忘仍是神经网络在顺序训练任务时的核心挑战。本文将持续学习重新构想为一个控制问题，其中学习信号与保留信号在神经活动动态中相互竞争。我们将正则化惩罚转化为保留信号，以保护先前任务的表征。学习过程通过最小化整合新任务所需的控制努力来推进，同时与先前任务的保留需求竞争。在平衡状态下，神经活动产生的权重更新隐式编码了完整的先验任务曲率，这一特性称为持续自然梯度，无需显式存储曲率。实验表明，该学习框架能够恢复真实的先验任务曲率，并实现任务区分，在标准基准上优于现有方法，且无需回放。
原文摘要 · Abstract (English)
Catastrophic forgetting remains a fundamental challenge for neural networks when tasks are trained sequentially. In this work, we reformulate continual learning as a control problem where learning and preservation signals compete within neural activity dynamics. We convert regularization penalties into preservation signals that protect prior-task representations. Learning then proceeds by minimizing the control effort required to integrate new tasks while competing with the preservation of prior tasks. At equilibrium, the neural activities produce weight updates that implicitly encode the full prior-task curvature, a property we term the continual-natural gradient, requiring no explicit curvature storage. Experiments confirm that our learning framework recovers true prior-task curvature and enables task discrimination, outperforming existing methods on standard benchmarks without replay.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。