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:1809.03124 · Proceedings of the National Academy of Sciences

Approaching the adiabatic timescale with machine learning

B. M. Henson, D. K. Shin, Kieran F. Thomas, Jacob A. Ross, Michael Hush, Sean S. Hodgman, Andrew G. Truscott

Published Dec 7, 2018• 34 citations
arXivHTMLPDF

Abstract

The control and manipulation of quantum systems without excitation are challenging, due to the complexities in fully modeling such systems accurately and the difficulties in controlling these inherently fragile systems experimentally. For example, while protocols to decompress Bose-Einstein condensates (BECs) faster than the adiabatic timescale (without excitation or loss) have been well developed theoretically, experimental implementations of these protocols have yet to reach speeds faster than the adiabatic timescale. In this work, we experimentally demonstrate an alternative approach based on a machine-learning algorithm which makes progress toward this goal. The algorithm is given control of the coupled decompression and transport of a metastable helium condensate, with its performance determined after each experimental iteration by measuring the excitations of the resultant BEC. After each iteration the algorithm adjusts its internal model of the system to create an improved control output for the next iteration. Given sufficient control over the decompression, the algorithm converges to a solution that sets the current speed record in relation to the adiabatic timescale, beating out other experimental realizations based on theoretical approaches. This method presents a feasible approach for implementing fast-state preparations or transformations in other quantum systems, without requiring a solution to a theoretical model of the system. Implications for fundamental physics and cooling are discussed.

DOI
10.1073/pnas.1811501115

Related papers

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

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

Shortcuts to adiabaticity for trapped ultracold gases

Jean-François Schaff, P. Capuzzi, G. Labeyrie +1 more

We study experimentally and theoretically the controlled transfer of harmonically trapped ultracold gases between different quantum states. In particular, we experimentally demonstrate a fast decompression and displacement of both a non-interacting gas and an interacting Bose–Einstein condensate, which are initially at equilibrium. The decompression parameters are engineered such that the final state is identical to that obtained after a perfectly adiabatic transformation despite the fact that the fast decompression is performed in the strongly non-adiabatic regime. During the transfer the atomic sample goes through strongly out-of-equilibrium states, while the external confinement is modified until the system reaches the desired stationary state. The scheme is theoretically based on the invariants of motion and scaling equation techniques and can be generalized to decompression trajectories including an arbitrary deformation of the trap. It is also directly applicable to arbitrary initial non-equilibrium states.

arXiv:1105.2119•New Journal of Physics•Nov 11, 2011•105 citations

Faster State Preparation across Quantum Phase Transition Assisted by Reinforcement Learning

Shuai-Feng Guo, Feng Chen, Qi Liu +6 more

An energy gap develops near quantum critical point of quantum phase transition in a finite many-body (MB) system, facilitating the ground state transformation by adiabatic parameter change. In real application scenarios, however, the efficacy for such a protocol is compromised by the need to balance finite system lifetime with adiabaticity, as exemplified in a recent experiment that prepares three-mode balanced Dicke state near deterministically [Y.-Q. Zou et al., Proc. Natl. Acad. Sci. U.S.A. 115, 6381 (2018)PNASA60027-842410.1073/pnas.1715105115]. Instead of tracking the instantaneous ground state as unanimously required for most adiabatic crossing, this work reports a faster sweeping policy taking advantage of excited level dynamics. It is obtained based on deep reinforcement learning (DRL) from a multistep training scheme we develop. In the absence of loss, a fidelity ≥99% between prepared and the target Dicke state is achieved over a small fraction of the adiabatically required time. When loss is included, training is carried out according to an operational benchmark, the interferometric sensitivity of the prepared state instead of fidelity, leading to better sensitivity in about half of the previously reported time. Implemented in a Bose-Einstein condensate of ∼10^{4} ^{87}Rb atoms, the balanced three-mode Dicke state exhibiting an improved number squeezing of 13.02±0.20 dB is observed within 766 ms, highlighting the potential of DRL for quantum dynamics control and quantum state preparation in interacting MB systems.

arXiv:2011.11987•Physical Review Letters•Feb 9, 2021•43 citations

Automated optimization, learning and control for cold atom platforms

Martino Calzavara

Quantum technologies based on cold atom platforms, while benefiting from highly stable and scalable atomic systems, still rely on the precise characterization, calibration, and control of complex experimental apparatuses. In this thesis, we develop automated approaches to address these critical tasks, integrating physical insight with optimization and learning methods, while complying with operational constraints. The bulk of the work presented here consists of three research papers. In the first one, we introduce physics-inspired machine-learning models to optimize optical dipole potentials for shaping ultracold gasses, achieving order-of-magnitude improvements in speed. In the second paper, we investigate atom transport between optical tweezers, a key auxiliary operation for scalable quantum processors, and demonstrate how shortcuts to adiabaticity and quantum optimal control can minimize heating and excitations during fast transport. The third and last paper analyzes a class of optimization landscapes relevant for quantum control, deriving from first principles a set of classical surrogates. We show how time and energy constraints translate into limited bandwidth and derivatives for the landscape, with consequences for the design of regression models and for optimization, which we relate to bounds in associated metrics. Together, these results provide concrete examples of how automated optimization, learning, and control can expand the toolset available for building quantum firmware for cold atom platforms. The methods presented not only improve experimental performance but keep implementation overhead at a minimum, thereby simplifying operations and facilitating scale-up efforts.

•Jan 1, 2025•0 citations

Critical States Preparation With Deep Reinforcement Learning

Jia-Wen Yu, Yi-Ming Yu, Ke-Xiong Yan +4 more

The fast and efficient preparation of quantum critical states is a challenging yet crucial task for various quantum technologies. This difficulty is most particularly for systems near a quantum phase transition, where the closure of the energy gap fundamentally limits the timescale of adiabatic processes and thus precludes rapid state preparation. We propose a framework using deep reinforcement learning (DRL) to rapidly prepare quantum critical states, with broad extendibility to light-matter interaction systems. Specifically, a DRL agent optimizes a set of time-dependent control Hamiltonians to drive the system from an initial noncritical state to a target critical state within a finite time and over experimentally accessible parameter ranges. As a concrete application, we focus on the quantum Rabi model. The DRL-optimized time-dependent control Hamiltonian yield a final state with high-fidelity ($>0.999$) to the target critical state. The protocol can be readily extended to other quantum critical systems described by light-matter interaction models, such as quantum Dicke model. This investigation provides a powerful new framework for preparing and manipulating quantum critical states.

quant-pharXiv:2603.09135•Mar 11, 2026•0 citations
The control and manipulation of quantum systems without excitation are challenging, due to the complexities in fully modeling such systems accurately and the difficulties in controlling these inherently fragile systems experimentally. For example, while protocols to decompress Bose-Einstein condensates (BECs) faster than the adiabatic timescale (without excitation or loss) have been well developed theoretically, experimental implementations of these protocols have yet to reach speeds faster than the adiabatic timescale. In this work, we experimentally demonstrate an alternative approach based on a machine-learning algorithm which makes progress toward this goal. The algorithm is given control of the coupled decompression and transport of a metastable helium condensate, with its performance determined after each experimental iteration by measuring the excitations of the resultant BEC. After each iteration the algorithm adjusts its internal model of the system to create an improved control output for the next iteration. Given sufficient control over the decompression, the algorithm converges to a solution that sets the current speed record in relation to the adiabatic timescale, beating out other experimental realizations based on theoretical approaches. This method presents a feasible approach for implementing fast-state preparations or transformations in other quantum systems, without requiring a solution to a theoretical model of the system. Implications for fundamental physics and cooling are discussed.
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.
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.
We study experimentally and theoretically the controlled transfer of harmonically trapped ultracold gases between different quantum states. In particular, we experimentally demonstrate a fast decompression and displacement of both a non-interacting gas and an interacting Bose–Einstein condensate, which are initially at equilibrium. The decompression parameters are engineered such that the final state is identical to that obtained after a perfectly adiabatic transformation despite the fact that the fast decompression is performed in the strongly non-adiabatic regime. During the transfer the atomic sample goes through strongly out-of-equilibrium states, while the external confinement is modified until the system reaches the desired stationary state. The scheme is theoretically based on the invariants of motion and scaling equation techniques and can be generalized to decompression trajectories including an arbitrary deformation of the trap. It is also directly applicable to arbitrary initial non-equilibrium states.
An energy gap develops near quantum critical point of quantum phase transition in a finite many-body (MB) system, facilitating the ground state transformation by adiabatic parameter change. In real application scenarios, however, the efficacy for such a protocol is compromised by the need to balance finite system lifetime with adiabaticity, as exemplified in a recent experiment that prepares three-mode balanced Dicke state near deterministically [Y.-Q. Zou et al., Proc. Natl. Acad. Sci. U.S.A. 115, 6381 (2018)PNASA60027-842410.1073/pnas.1715105115]. Instead of tracking the instantaneous ground state as unanimously required for most adiabatic crossing, this work reports a faster sweeping policy taking advantage of excited level dynamics. It is obtained based on deep reinforcement learning (DRL) from a multistep training scheme we develop. In the absence of loss, a fidelity ≥99% between prepared and the target Dicke state is achieved over a small fraction of the adiabatically required time. When loss is included, training is carried out according to an operational benchmark, the interferometric sensitivity of the prepared state instead of fidelity, leading to better sensitivity in about half of the previously reported time. Implemented in a Bose-Einstein condensate of ∼10^{4} ^{87}Rb atoms, the balanced three-mode Dicke state exhibiting an improved number squeezing of 13.02±0.20 dB is observed within 766 ms, highlighting the potential of DRL for quantum dynamics control and quantum state preparation in interacting MB systems.
Quantum technologies based on cold atom platforms, while benefiting from highly stable and scalable atomic systems, still rely on the precise characterization, calibration, and control of complex experimental apparatuses. In this thesis, we develop automated approaches to address these critical tasks, integrating physical insight with optimization and learning methods, while complying with operational constraints. The bulk of the work presented here consists of three research papers. In the first one, we introduce physics-inspired machine-learning models to optimize optical dipole potentials for shaping ultracold gasses, achieving order-of-magnitude improvements in speed. In the second paper, we investigate atom transport between optical tweezers, a key auxiliary operation for scalable quantum processors, and demonstrate how shortcuts to adiabaticity and quantum optimal control can minimize heating and excitations during fast transport. The third and last paper analyzes a class of optimization landscapes relevant for quantum control, deriving from first principles a set of classical surrogates. We show how time and energy constraints translate into limited bandwidth and derivatives for the landscape, with consequences for the design of regression models and for optimization, which we relate to bounds in associated metrics. Together, these results provide concrete examples of how automated optimization, learning, and control can expand the toolset available for building quantum firmware for cold atom platforms. The methods presented not only improve experimental performance but keep implementation overhead at a minimum, thereby simplifying operations and facilitating scale-up efforts.
The fast and efficient preparation of quantum critical states is a challenging yet crucial task for various quantum technologies. This difficulty is most particularly for systems near a quantum phase transition, where the closure of the energy gap fundamentally limits the timescale of adiabatic processes and thus precludes rapid state preparation. We propose a framework using deep reinforcement learning (DRL) to rapidly prepare quantum critical states, with broad extendibility to light-matter interaction systems. Specifically, a DRL agent optimizes a set of time-dependent control Hamiltonians to drive the system from an initial noncritical state to a target critical state within a finite time and over experimentally accessible parameter ranges. As a concrete application, we focus on the quantum Rabi model. The DRL-optimized time-dependent control Hamiltonian yield a final state with high-fidelity (>0.999>0.999>0.999) to the target critical state. The protocol can be readily extended to other quantum critical systems described by light-matter interaction models, such as quantum Dicke model. This investigation provides a powerful new framework for preparing and manipulating quantum critical states.