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:2107.00933 · Journal of Magnetic Resonance

Optimal control gradient precision trade-offs: Application to fast generation of DeepControl libraries for MRI

Mads Sloth Vinding, David L. Goodwin, Ilya Kuprov, Torben E. Lund

Published Oct 27, 2021• 11 citations
arXivHTMLPDF

Abstract

We have recently demonstrated supervised deep learning methods for rapid generation of radiofrequency pulses in magnetic resonance imaging (https://doi.org/10.1002/mrm.27740, https://doi.org/10.1002/mrm.28667). Unlike the previous iterative optimization approaches, deep learning methods generate a pulse using a fixed number of floating-point operations - this is important in MRI, where patient-specific pulses preferably must be produced in real time. However, deep learning requires vast training libraries, which must be generated using the traditional methods, e.g., iterative quantum optimal control methods. Those methods are usually variations of gradient descent, and the calculation of the gradient of the performance metric with respect to the pulse waveform can be the most numerically intensive step. In this communication, we explore various ways in which the calculation of gradients in quantum optimal control theory may be accelerated. Four optimization avenues are explored: truncated commutator series expansions at zeroth and first order, a novel midpoint truncation scheme at first order, and the exact complex-step method. For the spin systems relevant to MRI, the first-order midpoint truncation is found to be sufficiently accurate, but also significantly faster than the machine precision gradient. This makes the generation of training databases for the machine learning methods considerably more realistic.

DOI
10.1016/j.jmr.2021.107094

Related papers

基于深度学习的量子控制优化

Jincheng Zhang

This paper presents a novel approach to quantum control optimization utilizing deep learning techniques. Traditional quantum control optimization methods often rely on complex analytical calculations or computationally intensive numerical simulations, limiting their applicability to complex quantum systems. This work proposes a framework where a deep learning model is trained to directly learn optimal control pulses for a given quantum system. The model, trained on simulated quantum dynamics, effectively learns the mapping between system states and control parameters. This automated approach offers the potential to significantly accelerate the process of quantum control design and potentially enable control of systems beyond the reach of conventional optimization methods. The core claim of this work is the automation of quantum control through deep learning. The central mechanism involves the construction of a deep learning model to simulate quantum behavior and learn optimal control pulses. The resulting system demonstrates a promising path towards efficient and adaptive quantum control.

•Sep 3, 2026•0 citations

Approximate dynamics leading to more optimal control: Efficient exact derivatives

Jesper Hasseriis Mohr Jensen, Frederik Møller, Jens Jakob Sørensen +1 more

Accurate derivatives are important for locally traversing and converging efficiently in quantum optimization landscapes. By deriving analytically exact control derivatives (gradient and Hessian) for unitary control tasks, we show here that the computational feasibility of meeting this accuracy requirement depends on the choice of propagation scheme and problem representation. Even when exact propagation is sufficiently cheap it is, perhaps surprisingly, much more efficient to optimize the (appropriately) approximate propagators: Approximations in the dynamics are traded off for significant complexity reductions in the exact derivative calculations. Importantly, past the initial analytical considerations, only standard numerical techniques are explicitly required with straightforward application to realistic systems. These results are numerically verified for two concrete problems of increasing Hilbert space dimensionality. The best schemes obtain unit fidelity to machine precision, whereas the results for other schemes are separated consistently by orders of magnitude in computation time and in worst-case ten orders of magnitude in achievable fidelity. Since these gaps continually increase with system size and complexity, this methodology allows numerically efficient optimization of very high-dimensional dynamics, e.g., in many-body contexts, operating in the high-fidelity regime.

arXiv:2005.09943•Physical Review A•Jun 21, 2021•19 citations

Improving quantum linear system solvers via a gradient descent perspective

Sander Gribling, Iordanis Kerenidis, Dániel Szilágyi

Solving systems of linear equations is one of the most important primitives in quantum computing that has the potential to provide a practical quantum advantage in many different areas, including in optimization, simulation, and machine learning. In this work, we revisit quantum linear system solvers from the perspective of convex optimization, and in particular gradient descent-type algorithms. This leads to a considerable constant-factor improvement in the runtime (or, conversely, a several orders of magnitude smaller error with the same runtime/circuit depth). More precisely, we first show how the asymptotically optimal quantum linear system solver of Childs, Kothari, and Somma is related to the gradient descent algorithm on the convex function $\|A\vec x - \vec b\|_2^2$: their linear system solver is based on a truncation in the Chebyshev basis of the degree-$(t-1)$ polynomial (in $A$) that maps the initial solution $\vec{x}_1 := \vec{b}$ to the $t$-th iterate $\vec{x}_t$ in the basic gradient descent algorithm. Then, instead of starting from the basic gradient descent algorithm, we use the optimal Chebyshev iteration method (which can be viewed as an accelerated gradient descent algorithm) and show that this leads to considerable improvements in the quantum solver.

arXiv:2109.04248•Sep 9, 2021•1 citation

##基于深度学习的量子态优化控制

Jincheng Zhang

This paper explores the application of deep learning techniques to the field of quantum state optimization control. Traditional methods for optimizing control pulses for quantum systems are often computationally intensive and may struggle to handle the complexity of realistic quantum systems. This research proposes a novel framework utilizing deep learning to automatically optimize control pulses, aiming for precise control and state evolution of quantum systems. The core of the approach involves constructing a deep learning model where the quantum system state is the input and the control pulse is the output. This model is trained using either reinforcement learning or supervised learning algorithms. The potential of this method to significantly accelerate the optimization process and facilitate control of complex quantum systems is investigated. The resulting optimized control pulses demonstrate improved performance compared to conventional optimization strategies. This work lays the foundation for automated quantum control and opens new possibilities for manipulating quantum systems with greater precision and efficiency.

•Sep 5, 2026•0 citations

Optimizing Quantum Variational Circuits with Deep Reinforcement Learning

Owen Lockwood

Quantum Machine Learning (QML) is considered to be one of the most promising applications of near term quantum devices. However, the optimization of quantum machine learning models presents numerous challenges arising from the imperfections of hardware and the fundamental obstacles in navigating an exponentially scaling Hilbert space. In this work, we evaluate the potential of contemporary methods in deep reinforcement learning to augment gradient based optimization routines in quantum variational circuits. We find that reinforcement learning augmented optimizers consistently outperform gradient descent in noisy environments. All code and pretrained weights are available to replicate the results or deploy the models at: https://github.com/lockwo/rl_qvc_opt.

arXiv:2109.03188•Sep 7, 2021•2 citations

Sparse Probabilistic Synthesis of Quantum Operations

Bálint Koczor

Successful implementations of quantum technologies require protocols and algorithms that use as few quantum resources as possible. However, many important quantum operations, such as continuous rotation gates in quantum computing or broadband pulses in NMR or magnetic resonance imaging (MRI) applications, can only be implemented approximately using finite quantum resources. This work develops an approach that, on average—at the cost of a modestly increased measurement repetition rate—enables exact implementations. One proceeds by first building a library of a large number of different approximations to the desired gate operation; by randomly selecting these operations according to a preoptimized probability distribution, one can on average implement the desired operation with a rigorously controllable approximation error. The approach relies on sophisticated tools from convex optimization to efficiently find optimal probability distributions. A diverse spectrum of applications are demonstrated as (a) exactly synthesizing rotations in fault-tolerant quantum computers using only low T -count circuits and (b) synthesizing broadband and band-selective pulses of superior performance in quantum optimal control with (c) further applications in NMR or MRI. The approach is very general and a broad spectrum of practical applications in quantum technologies are explicitly demonstrated. Published by the American Physical Society 2024

arXiv:2402.15550•PRX Quantum•Dec 31, 2024•6 citations
We have recently demonstrated supervised deep learning methods for rapid generation of radiofrequency pulses in magnetic resonance imaging (https://doi.org/10.1002/mrm.27740, https://doi.org/10.1002/mrm.28667). Unlike the previous iterative optimization approaches, deep learning methods generate a pulse using a fixed number of floating-point operations - this is important in MRI, where patient-specific pulses preferably must be produced in real time. However, deep learning requires vast training libraries, which must be generated using the traditional methods, e.g., iterative quantum optimal control methods. Those methods are usually variations of gradient descent, and the calculation of the gradient of the performance metric with respect to the pulse waveform can be the most numerically intensive step. In this communication, we explore various ways in which the calculation of gradients in quantum optimal control theory may be accelerated. Four optimization avenues are explored: truncated commutator series expansions at zeroth and first order, a novel midpoint truncation scheme at first order, and the exact complex-step method. For the spin systems relevant to MRI, the first-order midpoint truncation is found to be sufficiently accurate, but also significantly faster than the machine precision gradient. This makes the generation of training databases for the machine learning methods considerably more realistic.
This paper presents a novel approach to quantum control optimization utilizing deep learning techniques. Traditional quantum control optimization methods often rely on complex analytical calculations or computationally intensive numerical simulations, limiting their applicability to complex quantum systems. This work proposes a framework where a deep learning model is trained to directly learn optimal control pulses for a given quantum system. The model, trained on simulated quantum dynamics, effectively learns the mapping between system states and control parameters. This automated approach offers the potential to significantly accelerate the process of quantum control design and potentially enable control of systems beyond the reach of conventional optimization methods. The core claim of this work is the automation of quantum control through deep learning. The central mechanism involves the construction of a deep learning model to simulate quantum behavior and learn optimal control pulses. The resulting system demonstrates a promising path towards efficient and adaptive quantum control.
Accurate derivatives are important for locally traversing and converging efficiently in quantum optimization landscapes. By deriving analytically exact control derivatives (gradient and Hessian) for unitary control tasks, we show here that the computational feasibility of meeting this accuracy requirement depends on the choice of propagation scheme and problem representation. Even when exact propagation is sufficiently cheap it is, perhaps surprisingly, much more efficient to optimize the (appropriately) approximate propagators: Approximations in the dynamics are traded off for significant complexity reductions in the exact derivative calculations. Importantly, past the initial analytical considerations, only standard numerical techniques are explicitly required with straightforward application to realistic systems. These results are numerically verified for two concrete problems of increasing Hilbert space dimensionality. The best schemes obtain unit fidelity to machine precision, whereas the results for other schemes are separated consistently by orders of magnitude in computation time and in worst-case ten orders of magnitude in achievable fidelity. Since these gaps continually increase with system size and complexity, this methodology allows numerically efficient optimization of very high-dimensional dynamics, e.g., in many-body contexts, operating in the high-fidelity regime.
Solving systems of linear equations is one of the most important primitives in quantum computing that has the potential to provide a practical quantum advantage in many different areas, including in optimization, simulation, and machine learning. In this work, we revisit quantum linear system solvers from the perspective of convex optimization, and in particular gradient descent-type algorithms. This leads to a considerable constant-factor improvement in the runtime (or, conversely, a several orders of magnitude smaller error with the same runtime/circuit depth). More precisely, we first show how the asymptotically optimal quantum linear system solver of Childs, Kothari, and Somma is related to the gradient descent algorithm on the convex function ∥Ax⃗−b⃗∥22\|A\vec x - \vec b\|_2^2∥Ax−b∥22​: their linear system solver is based on a truncation in the Chebyshev basis of the degree-(t−1)(t-1)(t−1) polynomial (in AAA) that maps the initial solution x⃗1:=b⃗\vec{x}_1 := \vec{b}x1​:=b to the ttt-th iterate x⃗t\vec{x}_txt​ in the basic gradient descent algorithm. Then, instead of starting from the basic gradient descent algorithm, we use the optimal Chebyshev iteration method (which can be viewed as an accelerated gradient descent algorithm) and show that this leads to considerable improvements in the quantum solver.
This paper explores the application of deep learning techniques to the field of quantum state optimization control. Traditional methods for optimizing control pulses for quantum systems are often computationally intensive and may struggle to handle the complexity of realistic quantum systems. This research proposes a novel framework utilizing deep learning to automatically optimize control pulses, aiming for precise control and state evolution of quantum systems. The core of the approach involves constructing a deep learning model where the quantum system state is the input and the control pulse is the output. This model is trained using either reinforcement learning or supervised learning algorithms. The potential of this method to significantly accelerate the optimization process and facilitate control of complex quantum systems is investigated. The resulting optimized control pulses demonstrate improved performance compared to conventional optimization strategies. This work lays the foundation for automated quantum control and opens new possibilities for manipulating quantum systems with greater precision and efficiency.
Quantum Machine Learning (QML) is considered to be one of the most promising applications of near term quantum devices. However, the optimization of quantum machine learning models presents numerous challenges arising from the imperfections of hardware and the fundamental obstacles in navigating an exponentially scaling Hilbert space. In this work, we evaluate the potential of contemporary methods in deep reinforcement learning to augment gradient based optimization routines in quantum variational circuits. We find that reinforcement learning augmented optimizers consistently outperform gradient descent in noisy environments. All code and pretrained weights are available to replicate the results or deploy the models at: https://github.com/lockwo/rl_qvc_opt.
Successful implementations of quantum technologies require protocols and algorithms that use as few quantum resources as possible. However, many important quantum operations, such as continuous rotation gates in quantum computing or broadband pulses in NMR or magnetic resonance imaging (MRI) applications, can only be implemented approximately using finite quantum resources. This work develops an approach that, on average—at the cost of a modestly increased measurement repetition rate—enables exact implementations. One proceeds by first building a library of a large number of different approximations to the desired gate operation; by randomly selecting these operations according to a preoptimized probability distribution, one can on average implement the desired operation with a rigorously controllable approximation error. The approach relies on sophisticated tools from convex optimization to efficiently find optimal probability distributions. A diverse spectrum of applications are demonstrated as (a) exactly synthesizing rotations in fault-tolerant quantum computers using only low T -count circuits and (b) synthesizing broadband and band-selective pulses of superior performance in quantum optimal control with (c) further applications in NMR or MRI. The approach is very general and a broad spectrum of practical applications in quantum technologies are explicitly demonstrated. Published by the American Physical Society 2024