Metalearning with Hebbian Fast Weights

TL;DR

This paper introduces a novel meta-learning model that combines slow weights learned via SGD with fast Hebbian weights for one-shot learning, achieving state-of-the-art results on multiple benchmarks.

Contribution

It unifies neural and associative memory approaches, enabling joint data representation and label binding in a single model for one-shot learning.

Findings

Achieves state-of-the-art results on Omniglot, Mini-ImageNet, and Penn Treebank.

Effectively combines slow and fast weights for meta-learning.

Demonstrates the effectiveness of Hebbian learning in neural models.

Abstract

We unify recent neural approaches to one-shot learning with older ideas of associative memory in a model for metalearning. Our model learns jointly to represent data and to bind class labels to representations in a single shot. It builds representations via slow weights, learned across tasks through SGD, while fast weights constructed by a Hebbian learning rule implement one-shot binding for each new task. On the Omniglot, Mini-ImageNet, and Penn Treebank one-shot learning benchmarks, our model achieves state-of-the-art results.