Towards meaningful physics from generative models

Abstract

In several physical systems, important properties characterizing the system itself are theoretically related with specific degrees of freedom. Although standard Monte Carlo simulations provide an effective tool to accurately reconstruct the physical configurations of the system, they are unable to isolate the different contributions corresponding to different degrees of freedom. Here we show that unsupervised deep learning can become a valid support to MC simulation, coupling useful insights in the phases detection task with good reconstruction performance. As a testbed we consider the 2D XY model, showing that a deep neural network based on variational autoencoders can detect the continuous Kosterlitz-Thouless (KT) transitions, and that, if endowed with the appropriate constrains, they generate configurations with meaningful physical content.

References in corpus (16)

• Auto-Encoding Variational Bayes
• Stochastic Backpropagation and Approximate Inference in Deep Generative Models
• Machine learning phases of matter
• Tutorial on Variational Autoencoders
• Learning phase transitions by confusion
• Discovering Phase Transitions with Unsupervised Learning
• Unsupervised learning of phase transitions: from principal component analysis to variational autoencoders
• Discovering Phases, Phase Transitions and Crossovers through Unsupervised Machine Learning: A critical examination
• Learning Thermodynamics with Boltzmann Machines
• Self-Learning Monte Carlo Method
• Accelerate Monte Carlo Simulations with Restricted Boltzmann Machines
• Detection of phase transition via convolutional neural network
• An exact mapping between the Variational Renormalization Group and Deep Learning
• Probing many-body localization with neural networks
• Restricted Boltzmann Machines for the Long Range Ising Models
• Sampling algorithms for validation of supervised learning models for Ising-like systems

Cited by in corpus (9)

• Machine learning and the physical sciences
• Asymptotically unbiased estimation of physical observables with neural samplers
• Machine learning of phase transitions in the percolation and XY models
• Parameter diagnostics of phases and phase transition learning by neural networks
• Self-learning Monte Carlo with Deep Neural Networks
• Measurement-induced criticality as a data-structure transition
• Efficient Modelling of Trivializing Maps for Lattice ${\varphi }^4$ Theory Using Normalizing Flows: A First Look at Scalability
• Revealing quantum chaos with machine learning
• AdvNF: Reducing Mode Collapse in Conditional Normalising Flows using Adversarial Learning