Detection of Berezinskii-Kosterlitz-Thouless transition via generative adversarial networks

Anno: 2022

Autori: Contessi Daniele; Ricci Elisa; Recati Alessio; Rizzi Matte

Affiliazione autori: Dipartimento di Fisica, Universita di Trento & INO-CNR BEC Center, Povo, 38123, Italy; Forschungszentrum Julich GmbH, Institute of Quantum Control, Peter Grunberg Institut (PGI-8), Julich, 52425, Germany; Institute for Theoretical Physics, University of Cologne, Koln, D-50937, Germany; Dipartimento di Ingegneria e Scienza dell?Informazione, Universita di Trento, Deep Visual Learning research group, Fondazione Bruno Kessler (FBK), Povo, 38123, Italy

Abstract: The detection of phase transitions in quantum many-body systems with lowest possible prior knowledge of their details is among the most rousing goals of the flourishing application of machine-learning techniques to physical questions. Here, we train a Generative Adversarial Network (GAN) with the Entanglement Spectrum of a system bipartition, as extracted by means of Matrix Product States ansatze. We are able to identify gapless-to-gapped phase transitions in different one-dimensional models by looking at the machine inability to reconstruct outsider data with respect to the training set. We foresee that GAN-based methods will become instrumental in anomaly detection schemes applied to the determination of phase-diagrams. Copyright D. Contessi et al

Giornale/Rivista: SCIPOST PHYSICS

Volume: 12 (3)      Da Pagina: 107-1  A: 107-17

Maggiori informazioni: We acknowledge support from the Deutsche Forschungsgemeinschaft (DFG) , project grant 277101999, within the CRC network TR 183 (subproject B01) , the Eu-ropean Union (PASQuanS, Grant No. 817482) , the Alexander von Humboldt Foundation, from Provincia Autonoma di Trento, from Q@TN (the joint lab between University of Trento, FBK-Fondazione Bruno Kessler, INFN-National Institute for Nuclear Physics and CNR-National Research Council) and from the Italian MIUR under the PRIN2017 project CEnTraL. The MPS simulations were run on the JURECA Cluster at the Forschungszentrum Julich, with a code based on a flexible Abelian Symmetric Tensor Networks Library, developed in collaboration with the group of S. Montangero (Padua) .
Parole chiavi: phase-transitions
DOI: 10.21468/SciPostPhys.12.3.107