Compressing Neural Networks using the Variational Information Bottleneck

TL;DR

This paper introduces a neural network compression method based on the variational information bottleneck that effectively prunes neurons, reducing model size and computation while maintaining performance.

Contribution

It applies the information bottleneck principle with a variational bound to improve neuron pruning, outperforming existing methods in compression efficiency.

Findings

Achieves state-of-the-art compression rates

Reduces redundancy between layers effectively

Provides natural sparse regularization without extra tuning

Abstract

Neural networks can be compressed to reduce memory and computational requirements, or to increase accuracy by facilitating the use of a larger base architecture. In this paper we focus on pruning individual neurons, which can simultaneously trim model size, FLOPs, and run-time memory. To improve upon the performance of existing compression algorithms we utilize the information bottleneck principle instantiated via a tractable variational bound. Minimization of this information theoretic bound reduces the redundancy between adjacent layers by aggregating useful information into a subset of neurons that can be preserved. In contrast, the activations of disposable neurons are shut off via an attractive form of sparse regularization that emerges naturally from this framework, providing tangible advantages over traditional sparsity penalties without contributing additional tuning parameters to the energy landscape. We demonstrate state-of-the-art compression rates across an array of datasets and network architectures.