Disentangled Information Bottleneck

TL;DR

This paper introduces Disentangled Information Bottleneck (DisenIB), a novel approach that achieves maximum compression without sacrificing prediction performance, improving generalization and robustness in supervised learning.

Contribution

DisenIB provides a new framework for supervised disentangling that ensures maximum compression while maintaining prediction accuracy, addressing optimization issues of the traditional IB method.

Findings

DisenIB achieves maximum compression without loss in prediction performance.

The method improves generalization and robustness to adversarial attacks.

DisenIB enhances out-of-distribution detection and supervised disentangling.

Abstract

The information bottleneck (IB) method is a technique for extracting information that is relevant for predicting the target random variable from the source random variable, which is typically implemented by optimizing the IB Lagrangian that balances the compression and prediction terms. However, the IB Lagrangian is hard to optimize, and multiple trials for tuning values of Lagrangian multiplier are required. Moreover, we show that the prediction performance strictly decreases as the compression gets stronger during optimizing the IB Lagrangian. In this paper, we implement the IB method from the perspective of supervised disentangling. Specifically, we introduce Disentangled Information Bottleneck (DisenIB) that is consistent on compressing source maximally without target prediction performance loss (maximum compression). Theoretical and experimental results demonstrate that our method is consistent on maximum compression, and performs well in terms of generalization, robustness to adversarial attack, out-of-distribution detection, and supervised disentangling.