On Information Plane Analyses of Neural Network Classifiers -- A Review
arXiv:2003.09671 · doi:10.1109/TNNLS.2021.3089037
Abstract
We review the current literature concerned with information plane analyses of neural network classifiers. While the underlying information bottleneck theory and the claim that information-theoretic compression is causally linked to generalization are plausible, empirical evidence was found to be both supporting and conflicting. We review this evidence together with a detailed analysis of how the respective information quantities were estimated. Our survey suggests that compression visualized in information planes is not necessarily information-theoretic, but is rather often compatible with geometric compression of the latent representations. This insight gives the information plane a renewed justification. Aside from this, we shed light on the problem of estimating mutual information in deterministic neural networks and its consequences. Specifically, we argue that even in feed-forward neural networks the data processing inequality need not hold for estimates of mutual information. Similarly, while a fitting phase, in which the mutual information between the latent representation and the target increases, is necessary (but not sufficient) for good classification performance, depending on the specifics of mutual information estimation such a fitting phase need not be visible in the information plane.
arXiv comment12 pages, 3 figures; accepted for publication in IEEE Transactions on Neural Networks and Learning Systems. (c) 2021 IEEE
References in corpus (21)
- Estimating Mutual Information
- Methods for Interpreting and Understanding Deep Neural Networks
- Neural Tangent Kernel: Convergence and Generalization in Neural Networks
- Opening the Black Box of Deep Neural Networks via Information
- On the importance of single directions for generalization
- Learning Representations for Neural Network-Based Classification Using the Information Bottleneck Principle
- Nonlinear Information Bottleneck
- On Variational Bounds of Mutual Information
- Entropy and mutual information in models of deep neural networks
- Estimating Mixture Entropy with Pairwise Distances
- A Survey on Neural Network Interpretability
- Understanding Neural Networks and Individual Neuron Importance via Information-Ordered Cumulative Ablation
- Information Plane Analysis of Deep Neural Networks via Matrix-Based Renyi's Entropy and Tensor Kernels
- Information flows of diverse autoencoders
- Information Bottleneck and its Applications in Deep Learning
- On the Information Plane of Autoencoders
- Layer-wise Learning of Stochastic Neural Networks with Information Bottleneck
- Information in Infinite Ensembles of Infinitely-Wide Neural Networks
- Adaptive Estimators Show Information Compression in Deep Neural Networks
- What Information Does a ResNet Compress?
- Understanding Learning Dynamics of Binary Neural Networks via Information Bottleneck