Mutual Information, Neural Networks and the Renormalization Group Item Preview Given a field theory action, we train a flow-based hierarchical deep generative neural network to reproduce the boundary field ensemble from uncorrelated bulk field fluctuations. Title: Mutual Information, Neural Networks and the Renormalization Group. [1704.06279] Mutual Information, Neural Networks and the Renormalization Group Abstract: Physical systems differring in their microscopic details often display strikingly similar behaviour when probed at macroscopic scales. We demonstrate RG flow and extract the Ising critical exponent. Here we introduce the neural network renormalization group as a universal approach to design generic EHM for interacting field theories. The model performs hierarchical change-of-variables transformations from the physical space to a latent space with reduced mutual information. We present a variational renormalization group (RG) approach based on a reversible generative model with hierarchical architecture. 02/08/2018 ∙ by Shuo-Hui Li, et al. The model can learn hierarchical transformations between physical variables and renormalized collective variables. Browse our catalogue of tasks and access state-of-the-art solutions. Mutual information, neural networks and the renormalization group. Authors: Maciej Koch-Janusz, Zohar Ringel (Submitted on 20 Apr 2017 , last revised 24 Sep 2018 (this version, v2)) Abstract: Physical systems differring in their microscopic details often display strikingly similar behaviour when probed at macroscopic scales. Get the latest machine learning methods with code. Authors: Maciej Koch-Janusz, Zohar Ringel (Submitted on 20 Apr 2017 (this version), latest version 24 Sep 2018 ) Abstract: Physical systems differing in their microscopic details often display strikingly similar behaviour when probed at low energies. Physical systems differing in their microscopic details often display strikingly similar behaviour when probed at macroscopic scales. We apply the algorithm to classical statistical physics problems in one and two dimensions. Conversely, the neural network Tip: you can also follow us on Twitter Title: Mutual Information, Neural Networks and the Renormalization Group. ∙ 0 ∙ share We present a variational renormalization group approach using deep generative model composed of bijectors. We introduce an artificial neural network based on a model-independent, information-theoretic characterization of a real-space RG procedure, which performs this task. Neural Network Renormalization Group.