动态增益调节缓解持续学习中的性能下降问题
Dynamic gain neuromodulation attenuates the stability gap under joint training
- 引入动态增益机制，实现快速适应与强记忆平衡
- 在多任务联合训练下显著减少任务切换时的性能下滑
- 适合研究持续学习中稳定性与可塑性权衡的学者
持续学习中的稳定性缺口表现为引入新任务时旧任务性能暂时下降，反映出快速适应与强记忆之间的不匹配。尽管动量SGD和Adam等优化器具有隐式多时间尺度行为，仍存在明显稳定性缺口。值得注意的是，这种缺口在理想联合训练设置下依然存在，因此需在此条件下研究其成因。受去甲肾上腺素神经调制在不确定性下短暂提升神经元增益的启发，本文提出一种动态增益缩放机制，作为双时间尺度优化方法：通过瞬时增大有效更新幅度并动态重参数化前向传播权重，从而实证缓解过渡期的曲率放大。在无任务先验的域增量与类别增量设置下，对MNIST、CIFAR及mini-ImageNet进行实验，结果表明该方法有效降低稳定性缺口，同时保持较高准确率，增强了任务切换时的鲁棒性。
原文摘要 · Abstract (English)
Recent work in continual learning has highlighted the stability gap -- a temporary performance drop on previously learned tasks when new ones are introduced. This phenomenon reflects a mismatch between rapid adaptation and strong retention at task boundaries, underscoring the need for optimization mechanisms that balance plasticity and stability over abrupt distribution changes. While optimizers such as momentum-SGD and Adam introduce implicit multi-timescale behavior, they still exhibit pronounced stability gaps. Importantly, these gaps persist even under ideal joint training, making it crucial to study them in this setting to isolate their causes from other sources of forgetting. Motivated by how noradrenergic (neuromodulatory) bursts transiently increase neuronal gain under uncertainty, we introduce a dynamic gain scaling mechanism as a two-timescale optimization technique that balances adaptation and retention by transiently increasing the effective update magnitude while dynamically reparameterizing the weights governing the forward pass, thereby empirically mitigating transition-induced curvature amplification. Across domain- and class-incremental MNIST, CIFAR, and mini-ImageNet benchmarks under task-agnostic joint training, dynamic gain scaling effectively attenuates stability gaps while maintaining competitive accuracy, improving robustness at task transitions.
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。