Abstract:Enabling a neural network to sequentially learn multiple tasks is of great significance for expanding the applicability of neural networks in real-world applications. However, artificial neural networks face the well-known problem of catastrophic forgetting. What is worse, the degradation of previously learned skills becomes more severe as the task sequence increases, known as the long-term catastrophic forgetting. It is due to two facts: first, as the model learns more tasks, the intersection of the low-error parameter subspace satisfying for these tasks becomes smaller or even does not exist; second, when the model learns a new task, the cumulative error keeps increasing as the model tries to protect the parameter configuration of previous tasks from interference. Inspired by the memory consolidation mechanism in mammalian brains with synaptic plasticity, we propose a confrontation mechanism in which Adversarial Neural Pruning and synaptic Consolidation (ANPyC) is used to overcome the long-term catastrophic forgetting issue. The neural pruning acts as long-term depression to prune task-irrelevant parameters, while the novel synaptic consolidation acts as long-term potentiation to strengthen task-relevant parameters. During the training, this confrontation achieves a balance in that only crucial parameters remain, and non-significant parameters are freed to learn subsequent tasks. ANPyC avoids forgetting important information and makes the model efficient to learn a large number of tasks. Specifically, the neural pruning iteratively relaxes the current task's parameter conditions to expand the common parameter subspace of the task; the synaptic consolidation strategy, which consists of a structure-aware parameter-importance measurement and an element-wise parameter updating strategy, decreases the cumulative error when learning new tasks. Our approach encourages the synapse to be sparse and polarized, which enables long-term learning and memory. ANPyC exhibits effectiveness and generalization on both image classification and generation tasks with multiple layer perceptron, convolutional neural networks, and generative adversarial networks, and variational autoencoder. The full source code is available at https://github.com/GeoX-Lab/ANPyC.

Reducing Catastrophic Forgetting in Modular Neural Networks by Dynamic Information Balancing

Overcoming Long-Term Catastrophic Forgetting Through Adversarial Neural Pruning and Synaptic Consolidation

Diffusion-based neuromodulation can eliminate catastrophic forgetting in simple neural networks

Avoiding Catastrophe: Active Dendrites Enable Multi-Task Learning in Dynamic Environments

Overcoming catastrophic forgetting in neural networks

Measuring Catastrophic Forgetting in Neural Networks

Rotate your Networks: Better Weight Consolidation and Less Catastrophic Forgetting

Attention-Based Structural-Plasticity

Learning to Modulate Random Weights: Neuromodulation-inspired Neural Networks For Efficient Continual Learning

Memory Recall: A Simple Neural Network Training Framework Against Catastrophic Forgetting

A Methodology-Oriented Study of Catastrophic Forgetting in Incremental Deep Neural Networks

Towards Redundancy-Free Sub-networks in Continual Learning

Reducing catastrophic forgetting when evolving neural networks

Understanding Catastrophic Forgetting and Remembering in Continual Learning with Optimal Relevance Mapping

SynapNet: A Complementary Learning System Inspired Algorithm With Real-Time Application in Multimodal Perception

Anatomy of Catastrophic Forgetting: Hidden Representations and Task Semantics

A brain-inspired algorithm that mitigates catastrophic forgetting of artificial and spiking neural networks with low computational cost

Wide Neural Networks Forget Less Catastrophically

Alleviating catastrophic forgetting using context-dependent gating and synaptic stabilization

Learn to Grow: A Continual Structure Learning Framework for Overcoming Catastrophic Forgetting

Task-Aware Information Routing from Common Representation Space in Lifelong Learning