Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

Bulletin of the American Physical Society

String patterns and machine learning: probing the Fermi-Hubbard model with cold atoms

Annabelle Bohrdt

Published Jun 4, 2020• 0 citations
OpenAlex

Related papers

Quantum Hamiltonian Learning for the Fermi-Hubbard Model

Hongkang Ni, Haoya Li, Lexing Ying

arXiv:2312.17390•Acta Applicandae Mathematicae•Apr 30, 2024•11 citations

String patterns in the doped Hubbard model

Christie S. Chiu, Geoffrey Ji, Annabelle Bohrdt +6 more

Understanding strongly correlated quantum many-body states is one of the most difficult challenges in modern physics. For example, there remain fundamental open questions on the phase diagram of the Hubbard model, which describes strongly correlated electrons in solids. In this work, we realize the Hubbard Hamiltonian and search for specific patterns within the individual images of many realizations of strongly correlated ultracold fermions in an optical lattice. Upon doping a cold-atom antiferromagnet, we find consistency with geometric strings, entities that may explain the relationship between hole motion and spin order, in both pattern-based and conventional observables. Our results demonstrate the potential for pattern recognition to provide key insights into cold-atom quantum many-body systems.

arXiv:1810.03584•Science•Jul 18, 2019•210 citations

Machine learning and the magnetic phases of correlated fermions

Ehsan Khatami, Kelvin Ch’ng, Nick Vasquez +2 more

Bulletin of the American Physical Society•Mar 5, 2018•0 citations

Machine learning of quantum many-fermion systems

Simon Trebst

Bulletin of the American Physical Society•Mar 5, 2018•0 citations

Machine Learning Detection of Correlations in Snapshots of Ultracold Atoms in Optical Lattices

Stephanie L. Striegel, Eduardo Ibarra-García-Padilla, Ehsan Khatami

Abstract Recent proposals have suggested the use of supervised learning with convolutional neural networks to shed light on some of the less well known phases of the Fermi-Hubbard model through the classification of snapshots from the quantum gas microscopy of ultracold atoms in optical lattices. However, there have been challenges in the interpretability of networks with more than one convolutional filter coupled to the input images. Here, we expand on previous work by considering multiple filters in the first convolutional layer and developing a process for analyzing the physical relevance of patterns obtained in the trained filters. We benchmark our approach at half-filling, where strong antiferromagnetic correlations are known to be present, and we find that upon hole doping, previously unknown patterns arise at temperatures below the tunneling amplitude. These patterns may be a signature of interesting arrangements of fermions in the lattice.

arXiv:2310.03267•Springer proceedings in physics•Jan 1, 2026•4 citations

Classifying Pattern Formation in Materials via Machine Learning

Lukasz Burzawa, Shuo Liu, E. W. Carlson

Scanning probe experiments such as scanning tunneling microscopy (STM) and atomic force microscopy (AFM) on strongly correlated materials often reveal complex pattern formation that occurs on multiple length scales. We have shown in two disparate correlated materials that the pattern formation is driven by proximity to a disorder-driven critical point. We developed new analysis concepts and techniques that relate the observed pattern formation to critical exponents by analyzing the geometry and statistics of clusters observed in these experiments and converting that information into critical exponents. Machine learning algorithms can be helpful correlating data from scanning probe experiments to theoretical models of pattern formation. We analyze the use of machine learning algorithms for the identification of critical behavior and universality underlying scanning probe data sets taken from correlated materials. This method has complementary strengths to the cluster analysis methods. The cluster techniques have a clear physical interpretation while machine learning algorithms require less expertise from the user and are faster to implement. The complementary nature of the two techniques further facilitates our understanding of correlated materials. The training of machine learning algorithms has been done through artificial neural networks. The neural net was trained using data from theoretical simulations of percolation and Ising models. The trained net had a 3.00% average classification error during testing. This proves that machine learning algorithms can successfully distinguish whether the complex pattern formation of a specific novel material is governed by uncorrelated percolation or by an interaction model fixed point.

•Jan 1, 2016•0 citations
Understanding strongly correlated quantum many-body states is one of the most difficult challenges in modern physics. For example, there remain fundamental open questions on the phase diagram of the Hubbard model, which describes strongly correlated electrons in solids. In this work, we realize the Hubbard Hamiltonian and search for specific patterns within the individual images of many realizations of strongly correlated ultracold fermions in an optical lattice. Upon doping a cold-atom antiferromagnet, we find consistency with geometric strings, entities that may explain the relationship between hole motion and spin order, in both pattern-based and conventional observables. Our results demonstrate the potential for pattern recognition to provide key insights into cold-atom quantum many-body systems.
Abstract Recent proposals have suggested the use of supervised learning with convolutional neural networks to shed light on some of the less well known phases of the Fermi-Hubbard model through the classification of snapshots from the quantum gas microscopy of ultracold atoms in optical lattices. However, there have been challenges in the interpretability of networks with more than one convolutional filter coupled to the input images. Here, we expand on previous work by considering multiple filters in the first convolutional layer and developing a process for analyzing the physical relevance of patterns obtained in the trained filters. We benchmark our approach at half-filling, where strong antiferromagnetic correlations are known to be present, and we find that upon hole doping, previously unknown patterns arise at temperatures below the tunneling amplitude. These patterns may be a signature of interesting arrangements of fermions in the lattice.
Scanning probe experiments such as scanning tunneling microscopy (STM) and atomic force microscopy (AFM) on strongly correlated materials often reveal complex pattern formation that occurs on multiple length scales. We have shown in two disparate correlated materials that the pattern formation is driven by proximity to a disorder-driven critical point. We developed new analysis concepts and techniques that relate the observed pattern formation to critical exponents by analyzing the geometry and statistics of clusters observed in these experiments and converting that information into critical exponents. Machine learning algorithms can be helpful correlating data from scanning probe experiments to theoretical models of pattern formation. We analyze the use of machine learning algorithms for the identification of critical behavior and universality underlying scanning probe data sets taken from correlated materials. This method has complementary strengths to the cluster analysis methods. The cluster techniques have a clear physical interpretation while machine learning algorithms require less expertise from the user and are faster to implement. The complementary nature of the two techniques further facilitates our understanding of correlated materials. The training of machine learning algorithms has been done through artificial neural networks. The neural net was trained using data from theoretical simulations of percolation and Ising models. The trained net had a 3.00% average classification error during testing. This proves that machine learning algorithms can successfully distinguish whether the complex pattern formation of a specific novel material is governed by uncorrelated percolation or by an interaction model fixed point.