Universal characteristics of deep neural network loss surfaces from random matrix theory

Abstract

This paper considers several aspects of random matrix universality in deep neural networks. Motivated by recent experimental work, we use universal properties of random matrices related to local statistics to derive practical implications for deep neural networks based on a realistic model of their Hessians. In particular we derive universal aspects of outliers in the spectra of deep neural networks and demonstrate the important role of random matrix local laws in popular pre-conditioning gradient descent algorithms. We also present insights into deep neural network loss surfaces from quite general arguments based on tools from statistical physics and random matrix theory.

42 pages

References in corpus (8)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• The Loss Surfaces of Multilayer Networks
• Explorations on high dimensional landscapes
• Measurements of Three-Level Hierarchical Structure in the Outliers in the Spectrum of Deepnet Hessians
• Spectrum of deformed random matrices and free probability
• Fluctuations in local quantum unique ergodicity for generalized Wigner matrices
• Random matrix analysis of deep neural network weight matrices
• More Than a Toy: Random Matrix Models Predict How Real-World Neural Representations Generalize