Transformer neural networks and quantum simulators: a hybrid approach for simulating strongly correlated systems

Owing to their great expressivity and versatility, neural networks have gained attention for simulating large two-dimensional quantum many-body systems. However, their expressivity comes with the cost of a challenging optimization due to the in general rugged and complicated loss landscape. Here, we...

Full description

Saved in:
Bibliographic Details
Main Authors: Hannah Lange, Guillaume Bornet, Gabriel Emperauger, Cheng Chen, Thierry Lahaye, Stefan Kienle, Antoine Browaeys, Annabelle Bohrdt
Format: Article
Language:English
Published: Verein zur Förderung des Open Access Publizierens in den Quantenwissenschaften 2025-03-01
Series:Quantum
Online Access:https://quantum-journal.org/papers/q-2025-03-26-1675/pdf/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850092005406801920
author Hannah Lange
Guillaume Bornet
Gabriel Emperauger
Cheng Chen
Thierry Lahaye
Stefan Kienle
Antoine Browaeys
Annabelle Bohrdt
author_facet Hannah Lange
Guillaume Bornet
Gabriel Emperauger
Cheng Chen
Thierry Lahaye
Stefan Kienle
Antoine Browaeys
Annabelle Bohrdt
author_sort Hannah Lange
collection DOAJ
description Owing to their great expressivity and versatility, neural networks have gained attention for simulating large two-dimensional quantum many-body systems. However, their expressivity comes with the cost of a challenging optimization due to the in general rugged and complicated loss landscape. Here, we present a hybrid optimization scheme for neural quantum states (NQS), involving a data-driven pretraining with numerical or experimental data and a second, Hamiltonian-driven optimization stage. By using both projective measurements from the computational basis as well as expectation values from other measurement configurations such as spin-spin correlations, our pretraining gives access to the sign structure of the state, yielding improved and faster convergence that is robust w.r.t. experimental imperfections and limited datasets. We apply the hybrid scheme to the ground state search for the 2D transverse field Ising model and dipolar XY model on $6\times 6$ and $10\times 10$ square lattices with a patched transformer wave function, using numerical data as well as experimental data from a programmable Rydberg quantum simulator [Chen et al., Nature 616 (2023)], and show that the information from a second measurement basis highly improves the performance. Our work paves the way for a reliable and efficient optimization of neural quantum states.
format Article
id doaj-art-308ceedd778642ce8227bec5fe1a856e
institution DOAJ
issn 2521-327X
language English
publishDate 2025-03-01
publisher Verein zur Förderung des Open Access Publizierens in den Quantenwissenschaften
record_format Article
series Quantum
spelling doaj-art-308ceedd778642ce8227bec5fe1a856e2025-08-20T02:42:14ZengVerein zur Förderung des Open Access Publizierens in den QuantenwissenschaftenQuantum2521-327X2025-03-019167510.22331/q-2025-03-26-167510.22331/q-2025-03-26-1675Transformer neural networks and quantum simulators: a hybrid approach for simulating strongly correlated systemsHannah LangeGuillaume BornetGabriel EmperaugerCheng ChenThierry LahayeStefan KienleAntoine BrowaeysAnnabelle BohrdtOwing to their great expressivity and versatility, neural networks have gained attention for simulating large two-dimensional quantum many-body systems. However, their expressivity comes with the cost of a challenging optimization due to the in general rugged and complicated loss landscape. Here, we present a hybrid optimization scheme for neural quantum states (NQS), involving a data-driven pretraining with numerical or experimental data and a second, Hamiltonian-driven optimization stage. By using both projective measurements from the computational basis as well as expectation values from other measurement configurations such as spin-spin correlations, our pretraining gives access to the sign structure of the state, yielding improved and faster convergence that is robust w.r.t. experimental imperfections and limited datasets. We apply the hybrid scheme to the ground state search for the 2D transverse field Ising model and dipolar XY model on $6\times 6$ and $10\times 10$ square lattices with a patched transformer wave function, using numerical data as well as experimental data from a programmable Rydberg quantum simulator [Chen et al., Nature 616 (2023)], and show that the information from a second measurement basis highly improves the performance. Our work paves the way for a reliable and efficient optimization of neural quantum states.https://quantum-journal.org/papers/q-2025-03-26-1675/pdf/
spellingShingle Hannah Lange
Guillaume Bornet
Gabriel Emperauger
Cheng Chen
Thierry Lahaye
Stefan Kienle
Antoine Browaeys
Annabelle Bohrdt
Transformer neural networks and quantum simulators: a hybrid approach for simulating strongly correlated systems
Quantum
title Transformer neural networks and quantum simulators: a hybrid approach for simulating strongly correlated systems
title_full Transformer neural networks and quantum simulators: a hybrid approach for simulating strongly correlated systems
title_fullStr Transformer neural networks and quantum simulators: a hybrid approach for simulating strongly correlated systems
title_full_unstemmed Transformer neural networks and quantum simulators: a hybrid approach for simulating strongly correlated systems
title_short Transformer neural networks and quantum simulators: a hybrid approach for simulating strongly correlated systems
title_sort transformer neural networks and quantum simulators a hybrid approach for simulating strongly correlated systems
url https://quantum-journal.org/papers/q-2025-03-26-1675/pdf/
work_keys_str_mv AT hannahlange transformerneuralnetworksandquantumsimulatorsahybridapproachforsimulatingstronglycorrelatedsystems
AT guillaumebornet transformerneuralnetworksandquantumsimulatorsahybridapproachforsimulatingstronglycorrelatedsystems
AT gabrielemperauger transformerneuralnetworksandquantumsimulatorsahybridapproachforsimulatingstronglycorrelatedsystems
AT chengchen transformerneuralnetworksandquantumsimulatorsahybridapproachforsimulatingstronglycorrelatedsystems
AT thierrylahaye transformerneuralnetworksandquantumsimulatorsahybridapproachforsimulatingstronglycorrelatedsystems
AT stefankienle transformerneuralnetworksandquantumsimulatorsahybridapproachforsimulatingstronglycorrelatedsystems
AT antoinebrowaeys transformerneuralnetworksandquantumsimulatorsahybridapproachforsimulatingstronglycorrelatedsystems
AT annabellebohrdt transformerneuralnetworksandquantumsimulatorsahybridapproachforsimulatingstronglycorrelatedsystems