Abstract
We investigate why deep neural networks suffer from \emph{loss of plasticity} in deep continual learning, failing to learn new tasks without reinitializing parameters. We show that this failure is preceded by Hessian spectral collapse at new-task initialization, where meaningful curvature directions vanish and gradient descent becomes ineffective. To characterize the necessary condition for successful training, we introduce the notion of $\tau$-trainability and show that current plasticity preserving algorithms can be unified under this framework. Targeting spectral collapse directly, we then discuss the Kronecker factored approximation of the Hessian, which motivates two regularization enhancements: maintaining high effective feature rank and applying $L2$ penalties. Experiments on continual supervised and reinforcement learning tasks confirm that combining these two regularizers effectively preserves plasticity.
Abstract (translated)
我们研究了深度神经网络在深层连续学习中为何会遭受\emph{可塑性丧失}，即无法在不重新初始化参数的情况下学习新任务。我们展示了这种失败是由新任务初始时Hessian谱坍缩所导致的，在此过程中有意义的曲率方向消失，梯度下降变得无效。为了描述成功训练的必要条件，我们引入了$\tau$-可训练性的概念，并表明当前保持可塑性的算法可以在这一框架下统一起来。针对直接解决谱坍缩的问题，接下来我们讨论Hessian的Kronecker因子近似方法，这促使了两种正则化增强策略：维持高有效特征秩和应用$L2$惩罚项。在连续监督学习和强化学习任务上的实验确认，结合这两种正则化器可以有效地保持可塑性。
URL
https://arxiv.org/abs/2509.22335
