Detection of Berezinskii-Kosterlitz-Thouless transition via Generative Adversarial Networks

Contessi, Daniele; Ricci, Elisa; Recati, Alessio; Rizzi, Matteo

doi:10.21468/SciPostPhys.12.3.107

SciPost Physics

Detection of Berezinskii-Kosterlitz-Thouless transition via Generative Adversarial Networks

Daniele Contessi, Elisa Ricci, Alessio Recati, Matteo Rizzi

SciPost Phys. 12, 107 (2022) · published 25 March 2022

doi: 10.21468/SciPostPhys.12.3.107
pdf
Submissions/Reports

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 ans\"atze. 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.

TY  - JOUR
PB  - SciPost Foundation
DO  - 10.21468/SciPostPhys.12.3.107
TI  - Detection of Berezinskii-Kosterlitz-Thouless transition via Generative Adversarial Networks
PY  - 2022/03/25
UR  - https://scipost.org/SciPostPhys.12.3.107
JF  - SciPost Physics
JA  - SciPost Phys.
VL  - 12
IS  - 3
SP  - 107
A1  - Contessi, Daniele
AU  - Ricci, Elisa
AU  - Recati, Alessio
AU  - Rizzi, Matteo
AB  - 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 ans\"atze. 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.
ER  -

@Article{10.21468/SciPostPhys.12.3.107,
	title={{Detection of Berezinskii-Kosterlitz-Thouless transition via Generative Adversarial Networks}},
	author={Daniele Contessi and Elisa Ricci and Alessio Recati and Matteo Rizzi},
	journal={SciPost Phys.},
	volume={12},
	pages={107},
	year={2022},
	publisher={SciPost},
	doi={10.21468/SciPostPhys.12.3.107},
	url={https://scipost.org/10.21468/SciPostPhys.12.3.107},
}

Cited by 4

Authors / Affiliations: mappings to Contributors and Organizations

See all Organizations.

¹ ² ³ Daniele Contessi,
³ ⁴ Elisa Ricci,
³ Alessio Recati,
¹ ² Matteo Rizzi

Funders for the research work leading to this publication

Alexander von Humboldt-Stiftung / Alexander von Humboldt Foundation
Deutsche Forschungsgemeinschaft / German Research FoundationDeutsche Forschungsgemeinschaft [DFG]
Horizon 2020 (through Organization: European Commission [EC])
Ministero dell’Istruzione, dell’Università e della Ricerca (MIUR) (through Organization: Ministero dell'Istruzione, dell'Università e della Ricerca / Ministry of Education, Universities and Research [MIUR])
Provincia Autonoma di Trento