Enabling Continual Learning with Differentiable Hebbian Plasticity

TL;DR

This paper introduces a Differentiable Hebbian Plasticity model that enhances continual learning by combining rapid episodic memory with slow-changing parameters, effectively reducing forgetting in neural networks across various benchmarks.

Contribution

The paper presents a novel Differentiable Hebbian Plasticity approach that integrates fast and slow learning components to improve continual learning performance.

Findings

Outperforms baseline methods on Permuted MNIST, Split MNIST, and Vision Datasets Mixture.

Effectively handles class imbalance and concept drift in the introduced imbalanced Permuted MNIST.

Requires no additional hyperparameters, simplifying implementation.

Abstract

Continual learning is the problem of sequentially learning new tasks or knowledge while protecting previously acquired knowledge. However, catastrophic forgetting poses a grand challenge for neural networks performing such learning process. Thus, neural networks that are deployed in the real world often struggle in scenarios where the data distribution is non-stationary (concept drift), imbalanced, or not always fully available, i.e., rare edge cases. We propose a Differentiable Hebbian Consolidation model which is composed of a Differentiable Hebbian Plasticity (DHP) Softmax layer that adds a rapid learning plastic component (compressed episodic memory) to the fixed (slow changing) parameters of the softmax output layer; enabling learned representations to be retained for a longer timescale. We demonstrate the flexibility of our method by integrating well-known task-specific synaptic consolidation methods to penalize changes in the slow weights that are important for each target task. We evaluate our approach on the Permuted MNIST, Split MNIST and Vision Datasets Mixture benchmarks, and introduce an imbalanced variant of Permuted MNIST -- a dataset that combines the challenges of class imbalance and concept drift. Our proposed model requires no additional hyperparameters and outperforms comparable baselines by reducing forgetting.