通过自适应线性注入保持持续学习中的可塑性
Preserving Plasticity in Continual Learning with Adaptive Linearity Injection
AI总结：
针对持续学习中的可塑性丧失问题，本文提出 AdaLin，通过可学习参数与梯度门控向神经元激活自适应注入线性，在多个分类、增量学习和离策略强化学习场景中显著提升性能。
摘要：
深度神经网络的可塑性丧失是指模型增量学习能力的逐渐下降，并已被确定为非平稳问题环境中学习的关键障碍。近期研究表明，深度线性网络往往对可塑性丧失具有较强的抵御能力。受此观察启发，我们提出了 Adaptive Linearization（AdaLin，自适应线性化），这是一种动态调整每个神经元激活函数以缓解可塑性丧失的通用方法。与依赖正则化或周期性重置的先前方法不同，AdaLin 为每个神经元配备一个可学习参数和一种门控机制，该机制根据神经元的梯度流向激活函数注入线性。这种自适应调制可确保充足的梯度信号并维持持续学习，同时不会引入额外超参数，也不需要显式任务边界。当与 ReLU、Tanh 和 GeLU 等传统激活函数一起使用时，我们证明 AdaLin 能够显著提升标准基准上的性能，包括 Random Label 和 Permuted MNIST、Random Label 和 Shuffled CIFAR-10，以及 Class-Split CIFAR-100。此外，其有效性还在更复杂的场景中得到展示，例如使用 ResNet-18 骨干网络在 CIFAR-100 上进行类别增量学习，以及缓解离策略强化学习智能体中的可塑性丧失。我们进行了一系列系统消融实验，结果表明神经元级自适应对良好性能至关重要，并分析了网络中可能与可塑性丧失相关的多项指标。
英文摘要：
Loss of plasticity in deep neural networks is the gradual reduction in a model's capacity to incrementally learn and has been identified as a key obstacle to learning in non-stationary problem settings. Recent work has shown that deep linear networks tend to be resilient towards loss of plasticity. Motivated by this observation, we propose Adaptive Linearization (AdaLin), a general approach that dynamically adapts each neuron's activation function to mitigate plasticity loss. Unlike prior methods that rely on regularization or periodic resets, AdaLin equips every neuron with a learnable parameter and a gating mechanism that injects linearity into the activation function based on its gradient flow. This adaptive modulation ensures sufficient gradient signal and sustains continual learning without introducing additional hyperparameters or requiring explicit task boundaries. When used with conventional activation functions like ReLU, Tanh, and GeLU, we demonstrate that AdaLin can significantly improve performance on standard benchmarks, including Random Label and Permuted MNIST, Random Label and Shuffled CIFAR-10, and Class-Split CIFAR-100. Furthermore, its efficacy is shown in more complex scenarios, such as class-incremental learning on CIFAR-100 with a ResNet-18 backbone, and in mitigating plasticity loss in off-policy reinforcement learning agents. We perform a systematic set of ablations that show that neuron-level adaptation is crucial for good performance and analyze a number of metrics in the network that might be correlated to loss of plasticity.
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。
