A robust estimator of mutual information for deep learning interpretability

Abstract

We develop the use of mutual information (MI), a well-established metric in information theory, to interpret the inner workings of deep learning models. To accurately estimate MI from a finite number of samples, we present GMM-MI (pronounced $‘‘$Jimmie$"$), an algorithm based on Gaussian mixture models that can be applied to both discrete and continuous settings. GMM-MI is computationally efficient, robust to the choice of hyperparameters and provides the uncertainty on the MI estimate due to the finite sample size. We extensively validate GMM-MI on toy data for which the ground truth MI is known, comparing its performance against established mutual information estimators. We then demonstrate the use of our MI estimator in the context of representation learning, working with synthetic data and physical datasets describing highly non-linear processes. We train deep learning models to encode high-dimensional data within a meaningful compressed (latent) representation, and use GMM-MI to quantify both the level of disentanglement between the latent variables, and their association with relevant physical quantities, thus unlocking the interpretability of the latent representation. We make GMM-MI publicly available.

arXiv comment30 pages, 8 figures. Minor changes to match version accepted for publication in Machine Learning: Science and Technology. GMM-MI available at https://github.com/dpiras/GMM-MI

References in corpus (26)

• A Universal Density Profile from Hierarchical Clustering
• The Structure of Cold Dark Matter Halos
• Estimating Mutual Information
• Evolution of structure in cold dark matter universes
• The Rise and Fall of Satellites in Galaxy Clusters
• How universal are the density profiles of dark halos?
• Extreme deconvolution: Inferring complete distribution functions from noisy, heterogeneous and incomplete observations
• Nonlinear Information Bottleneck
• Likelihood-free inference with neural compression of DES SV weak lensing map statistics
• Persistent Homology of Complex Networks for Dynamic State Detection
• Are mergers responsible for universal halo properties?
• Quantifying information transfer and mediation along causal pathways in complex systems
• Mutual Information as a Tool for Identifying Phase Transitions in Dynamical Complex Systems With Limited Data
• Estimation of mutual information for real-valued data with error bars and controlled bias
• Using mutual information to measure order in model glass-formers
• How much a galaxy knows about its large-scale environment?: An information theoretic perspective
• EmpiriciSN: Re-sampling Observed Supernova/Host Galaxy Populations using an XD Gaussian Mixture Model
• Filling the gaps: Gaussian mixture models from noisy, truncated or incomplete samples
• A study on the statistical significance of mutual information between morphology of a galaxy and its large-scale environment
• Symmetries and phase diagrams with real-space mutual information neural estimation
• Sufficiency of a Gaussian power spectrum likelihood for accurate cosmology from upcoming weak lensing surveys
• Machines Learn to Infer Stellar Parameters Just by Looking at a Large Number of Spectra
• Do galactic bars depend on environment?: An information theoretic analysis of Galaxy Zoo 2
• Can a conditioning on stellar mass explain the mutual information between morphology and environment?
• On the origin of red spirals: Does assembly bias play a role?
• Transfer entropy computation using the Perron-Frobenius operator

Cited by in corpus (9)

• Data Compression and Inference in Cosmology with Self-Supervised Machine Learning
• Explaining dark matter halo density profiles with neural networks
• Deep learning insights into cosmological structure formation
• Cosmological feedback from a halo assembly perspective
• A representation learning approach to probe for dynamical dark energy in matter power spectra
• Deep learning insights into non-universality in the halo mass function
• Data-Space Validation of High-Dimensional Models by Comparing Sample Quantiles
• $\Lambda$CDM and early dark energy in latent space: a data-driven parametrization of the CMB temperature power spectrum
• QUEST (Quasar Unsupervised Encoder and Synthesis Tool): A machine learning framework to generate quasar spectra