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

arXiv:2311.06795 · Physical Review A

Inspiration from machine learning on the example of optimization of the Bose-Einstein condensate of thulium atoms in a 1064-nm trap

D. A. Kumpilov, D. A. Pershin, I. S. Cojocaru, Vladimir Khlebnikov, I. A. Pyrkh, A. E. Rudnev, E. A. Fedotova, K. A. Khoruzhii, P. A. Aksentsev, D. V. Gaifutdinov, A. K. Zykova, V. V. Tsyganok, Alexey V. Akimov

Published Mar 19, 2024• 3 citations
arXivHTMLPDF

Abstract

The number of atoms in Bose-Einstein condensate determines the scale of experiments that can be performed, making it crucial for quantum simulations. Optimization of the number of atoms in the condensate is a complex problem which could be efficiently solved using the machine learning technique. Nevertheless, this approach usually does not give any insight in the underlying physics. Here we demonstrate the possibility to learn physics from machine learning on an example of condensation of thulium atoms at a 1064-nm dipole trap. Optimization of the number of condensed atoms revealed a saturation, which was explained as a limitation imposed by a three-body recombination process. This limitation was successfully overcome by leveraging Fano-Feshbach resonances.

DOI
10.1103/physreva.109.033313

Related papers

Machine learning for achieving Bose-Einstein condensation of thulium atoms

E. T. Davletov, V. V. Tsyganok, Vladimir Khlebnikov +3 more

Bose-Einstein condensation (BEC) is a powerful tool for a wide range of research activities, a large fraction of which is related to quantum simulations. Various problems may benefit from different atomic species, but cooling down novel species interesting for quantum simulations to BEC temperatures requires a substantial amount of optimization and is usually considered to be a difficult experimental task. In this work, we implemented the Bayesian machine learning technique to optimize the evaporative cooling of thulium atoms and achieved BEC in an optical dipole trap operating near 532 nm. The developed approach could be used to cool down other novel atomic species to quantum degeneracy without additional studies of their properties.

arXiv:2003.00346•Physical Review A•Jul 14, 2020•57 citations

Machine-learning-accelerated Bose-Einstein condensation

Zachary Vendeiro, Joshua Ramette, Alyssa Rudelis +5 more

Machine learning is emerging as a technology that can enhance physics experiment execution and data analysis. Here, we apply machine learning to accelerate the production of a Bose-Einstein condensate (BEC) of $^{87}\mathrm{Rb}$ atoms by Bayesian optimization of up to 55 control parameters. This approach enables us to prepare BECs of $2.8\ifmmode\times\else\texttimes\fi{}{10}^{3}$ optically trapped $^{87}\mathrm{Rb}$ atoms from a room-temperature gas in 575 ms. The algorithm achieves the fast BEC preparation by applying highly efficient Raman cooling to near quantum degeneracy, followed by a brief final evaporation. We anticipate that many other physics experiments with complex nonlinear system dynamics can be significantly enhanced by a similar machine-learning approach.

arXiv:2205.08057•Physical Review Research•Dec 27, 2022•35 citations

Can machine learning for quantum-gas experiments be explainable?

I. B. Spielman amd J. P. Zwolak

Virtually all aspects of many-body atomic physics are challenging: experiments are technically demanding, datasets have become enormous, and the memory and CPU requirements for classical simulation of generic quantum systems often scale exponentially with system size. Machine learning (ML) methods are already assisting in each of these areas and are poised to become transformative. Here, we focus on two specific applications of ML to cold-atom-based quantum simulators. These devices generally generate data in the form of images; we first showcase denoising of raw images and then identify solitonic waves in Bose-Einstein condensates. In both of these examples, we comment on the interplay between performance, model complexity, and interpretability.

cond-mat.quant-gasCross-listedarXiv:2605.18689•May 19, 2026•0 citations

Losses of thulium atoms from optical dipole traps operating at 532 and 1064 nm

V. V. Tsyganok, D. A. Pershin, Vladimir Khlebnikov +10 more

Recently, thulium has been condensed to Bose-Einstein condensate. Machine learning was used to avoid a detailed study of all obstacles, making cooling difficult. This paper analyzes the atomic loss mechanism for the 532-nm optical trap, used in the Bose-condensation experiment, and compares it with the alternative and more traditional micrometer-range optical dipole trap. We also measured the scalar and tensor polarizability of thulium at 1064 nm, which were found to be $167\ifmmode\pm\else\textpm\fi{}25$ a.u. ($275\ifmmode\pm\else\textpm\fi{}41\phantom{\rule{0.55542pt}{0ex}}\phantom{\rule{0.55542pt}{0ex}}\ifmmode\times\else\texttimes\fi{}{10}^{\ensuremath{-}41}\mathrm{F}\phantom{\rule{0.16em}{0ex}}{\mathrm{m}}^{2}$) and $\ensuremath{-}4\ifmmode\pm\else\textpm\fi{}1$ a.u. ($7\ifmmode\pm\else\textpm\fi{}2\phantom{\rule{0.55542pt}{0ex}}\phantom{\rule{0.55542pt}{0ex}}\ifmmode\times\else\texttimes\fi{}{10}^{\ensuremath{-}41}\phantom{\rule{0.28em}{0ex}}\mathrm{F}\phantom{\rule{0.16em}{0ex}}{\mathrm{m}}^{2}$), respectively.

arXiv:2302.02106•Physical Review A•Feb 21, 2023•9 citations

Bose-Einstein condensate as a diagnostic tool for an optical lattice formed by 1064-nm laser light

V. V. Tsyganok, D. A. Pershin, Vladimir Khlebnikov +10 more

Recently, thulium has been condensed to Bose-Einstein condensate. While the condensate of Thulium has a lot of applications in quantum simulations and other areas of physics, it can also serve as a unique diagnostic tool for many atomic experiments. In the present study, the Bose-Einstein condensate of Thulium was successfully loaded into an optical lattice and utilized to diagnose the lattice as well as detect unwanted reflections in the experiments with the 1064 nm optical lattice, which will further be used in a quantum gas microscope experiment.

arXiv:2303.10071•Physical Review A•Jul 11, 2023•4 citations

Fast machine-learning online optimization of ultra-cold-atom experiments

P. B. Wigley, P. J. Everitt, Anton van den Hengel +12 more

We apply an online optimization process based on machine learning to the production of Bose-Einstein condensates (BEC). BEC is typically created with an exponential evaporation ramp that is optimal for ergodic dynamics with two-body s-wave interactions and no other loss rates, but likely sub-optimal for real experiments. Through repeated machine-controlled scientific experimentation and observations our 'learner' discovers an optimal evaporation ramp for BEC production. In contrast to previous work, our learner uses a Gaussian process to develop a statistical model of the relationship between the parameters it controls and the quality of the BEC produced. We demonstrate that the Gaussian process machine learner is able to discover a ramp that produces high quality BECs in 10 times fewer iterations than a previously used online optimization technique. Furthermore, we show the internal model developed can be used to determine which parameters are essential in BEC creation and which are unimportant, providing insight into the optimization process of the system.

arXiv:1507.04964•Scientific Reports•May 16, 2016•226 citations
The number of atoms in Bose-Einstein condensate determines the scale of experiments that can be performed, making it crucial for quantum simulations. Optimization of the number of atoms in the condensate is a complex problem which could be efficiently solved using the machine learning technique. Nevertheless, this approach usually does not give any insight in the underlying physics. Here we demonstrate the possibility to learn physics from machine learning on an example of condensation of thulium atoms at a 1064-nm dipole trap. Optimization of the number of condensed atoms revealed a saturation, which was explained as a limitation imposed by a three-body recombination process. This limitation was successfully overcome by leveraging Fano-Feshbach resonances.
Bose-Einstein condensation (BEC) is a powerful tool for a wide range of research activities, a large fraction of which is related to quantum simulations. Various problems may benefit from different atomic species, but cooling down novel species interesting for quantum simulations to BEC temperatures requires a substantial amount of optimization and is usually considered to be a difficult experimental task. In this work, we implemented the Bayesian machine learning technique to optimize the evaporative cooling of thulium atoms and achieved BEC in an optical dipole trap operating near 532 nm. The developed approach could be used to cool down other novel atomic species to quantum degeneracy without additional studies of their properties.
Machine learning is emerging as a technology that can enhance physics experiment execution and data analysis. Here, we apply machine learning to accelerate the production of a Bose-Einstein condensate (BEC) of 87Rb^{87}\mathrm{Rb}87Rb atoms by Bayesian optimization of up to 55 control parameters. This approach enables us to prepare BECs of $2.8\ifmmode\times\else\texttimes\fi{}{10}^{3}$ optically trapped 87Rb^{87}\mathrm{Rb}87Rb atoms from a room-temperature gas in 575 ms. The algorithm achieves the fast BEC preparation by applying highly efficient Raman cooling to near quantum degeneracy, followed by a brief final evaporation. We anticipate that many other physics experiments with complex nonlinear system dynamics can be significantly enhanced by a similar machine-learning approach.
Virtually all aspects of many-body atomic physics are challenging: experiments are technically demanding, datasets have become enormous, and the memory and CPU requirements for classical simulation of generic quantum systems often scale exponentially with system size. Machine learning (ML) methods are already assisting in each of these areas and are poised to become transformative. Here, we focus on two specific applications of ML to cold-atom-based quantum simulators. These devices generally generate data in the form of images; we first showcase denoising of raw images and then identify solitonic waves in Bose-Einstein condensates. In both of these examples, we comment on the interplay between performance, model complexity, and interpretability.
Recently, thulium has been condensed to Bose-Einstein condensate. Machine learning was used to avoid a detailed study of all obstacles, making cooling difficult. This paper analyzes the atomic loss mechanism for the 532-nm optical trap, used in the Bose-condensation experiment, and compares it with the alternative and more traditional micrometer-range optical dipole trap. We also measured the scalar and tensor polarizability of thulium at 1064 nm, which were found to be $167\ifmmode\pm\else\textpm\fi{}25$ a.u. ($275\ifmmode\pm\else\textpm\fi{}41\phantom{\rule{0.55542pt}{0ex}}\phantom{\rule{0.55542pt}{0ex}}\ifmmode\times\else\texttimes\fi{}{10}^{\ensuremath{-}41}\mathrm{F}\phantom{\rule{0.16em}{0ex}}{\mathrm{m}}^{2}$) and $\ensuremath{-}4\ifmmode\pm\else\textpm\fi{}1$ a.u. ($7\ifmmode\pm\else\textpm\fi{}2\phantom{\rule{0.55542pt}{0ex}}\phantom{\rule{0.55542pt}{0ex}}\ifmmode\times\else\texttimes\fi{}{10}^{\ensuremath{-}41}\phantom{\rule{0.28em}{0ex}}\mathrm{F}\phantom{\rule{0.16em}{0ex}}{\mathrm{m}}^{2}$), respectively.
Recently, thulium has been condensed to Bose-Einstein condensate. While the condensate of Thulium has a lot of applications in quantum simulations and other areas of physics, it can also serve as a unique diagnostic tool for many atomic experiments. In the present study, the Bose-Einstein condensate of Thulium was successfully loaded into an optical lattice and utilized to diagnose the lattice as well as detect unwanted reflections in the experiments with the 1064 nm optical lattice, which will further be used in a quantum gas microscope experiment.
We apply an online optimization process based on machine learning to the production of Bose-Einstein condensates (BEC). BEC is typically created with an exponential evaporation ramp that is optimal for ergodic dynamics with two-body s-wave interactions and no other loss rates, but likely sub-optimal for real experiments. Through repeated machine-controlled scientific experimentation and observations our 'learner' discovers an optimal evaporation ramp for BEC production. In contrast to previous work, our learner uses a Gaussian process to develop a statistical model of the relationship between the parameters it controls and the quality of the BEC produced. We demonstrate that the Gaussian process machine learner is able to discover a ramp that produces high quality BECs in 10 times fewer iterations than a previously used online optimization technique. Furthermore, we show the internal model developed can be used to determine which parameters are essential in BEC creation and which are unimportant, providing insight into the optimization process of the system.