Reasoning-Modulated Representations

Petar Veličković, Matko Bošnjak, Thomas Kipf, Alexander Lerchner, raia hadsell, Razvan Pascanu, Charles Blundell

Keywords: representation learning, algorithmic reasoning, graph neural networks, relational learning

Abstract: Neural networks leverage robust internal representations in order to generalise. Learning them is difficult, and often requires a large training set that covers the data distribution densely. We study a common setting where our task is not purely opaque. Indeed, very often we may have access to information about the underlying system (e.g. that observations must obey certain laws of physics) that any "tabula rasa" neural network would need to re-learn from scratch, penalising data efficiency. We incorporate this information into a pre-trained reasoning module, and investigate its role in shaping the discovered representations in diverse self-supervised learning settings from pixels. Our approach paves the way for a new class of data-efficient representation learning.

One-sentence Summary: We investigate how the representations captured by neural networks can be modulated by abstract reasoning modules (which respect the dynamics of the underlying task), leading to gains in data efficiency.