SciPost Submission Page
Detection of Berezinskii-Kosterlitz-Thouless transition via Generative Adversarial Networks
by D. Contessi, E. Ricci, A. Recati, M. Rizzi
Submission summary
| Authors (as registered SciPost users): | Daniele Contessi |
| Submission information | |
|---|---|
| Preprint Link: | https://arxiv.org/abs/2110.05383v3 (pdf) |
| Code repository: | https://github.com/cerbero94/GAN_CP |
| Data repository: | https://github.com/cerbero94/GAN_CP/tree/main/data |
| Date accepted: | Feb. 28, 2022 |
| Date submitted: | Feb. 4, 2022, 7:04 p.m. |
| Submitted by: | Daniele Contessi |
| Submitted to: | SciPost Physics |
| Ontological classification | |
|---|---|
| Academic field: | Physics |
| Specialties: |
|
| Approaches: | Theoretical, Computational |
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.
Author comments upon resubmission
We directly replied point by point to their comments.
List of changes
Published as SciPost Phys. 12, 107 (2022)