Putting An End to End-to-End: Gradient-Isolated Learning of Representations

TL;DR

This paper introduces a biologically inspired, gradient-isolated training method for deep neural networks that enables scalable, label-free, self-supervised learning by training modules independently to preserve input information, achieving competitive results.

Contribution

It presents a novel module-based training approach that avoids end-to-end backpropagation, inspired by biological neural networks, allowing scalable, asynchronous training on unlabelled data.

Findings

Modules improve upon their predecessors' outputs

Achieves competitive downstream classification results

Enables large-scale distributed training

Abstract

We propose a novel deep learning method for local self-supervised representation learning that does not require labels nor end-to-end backpropagation but exploits the natural order in data instead. Inspired by the observation that biological neural networks appear to learn without backpropagating a global error signal, we split a deep neural network into a stack of gradient-isolated modules. Each module is trained to maximally preserve the information of its inputs using the InfoNCE bound from Oord et al. [2018]. Despite this greedy training, we demonstrate that each module improves upon the output of its predecessor, and that the representations created by the top module yield highly competitive results on downstream classification tasks in the audio and visual domain. The proposal enables optimizing modules asynchronously, allowing large-scale distributed training of very deep neural networks on unlabelled datasets.