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.

Neuroscience: Memory modification without catastrophe

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

Durable memories and efficient neural coding through mnemonic training using the method of loci

Attention-Based Structural-Plasticity

Overcoming catastrophic forgetting in neural networks

Alleviating catastrophic forgetting using context-dependent gating and synaptic stabilization

New frontiers in the study of memory mechanisms

Association: Remind Your GAN not to Forget

Behavioral Experiments for Understanding Catastrophic Forgetting

Changes in patterns of neural activity underlie a time-dependent transformation of memory in rats and humans

Catastrophic Importance of Catastrophic Forgetting

Contributions by metaplasticity to solving the Catastrophic Forgetting Problem

Natural forgetting reversibly modulates engram expression

Persistence of neuronal representations through time and damage in the hippocampus

Coordinating brain-distributed network activities in memory resistant to extinction

Memory Aware Synapses: Learning what (not) to forget

Measuring Catastrophic Forgetting in Neural Networks

The formation of an expanding memory representation in the hippocampus

Active Forgetting: Adaptation of Memory by Prefrontal Control

Temporary changes in large-scale memory neural networks after fear learning and extinction in healthy adults

Survival of the salient: Aversive learning rescues otherwise forgettable memories via neural reactivation and post-encoding hippocampal connectivity