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:2611.16365

Training quantum neural networks one job at a time

Muhammad Faryad, Rabab Alomairy

Published Nov 12, 2026• 0 citations
arXivHTMLPDF

Abstract

When a quantum processor is reached through a cloud queue, a variational algorithm pays for every round trip, and a single round trip can carry many circuits. The classical optimizer therefore decides how much is learned per job, not only per shot. We study this for a quantum neural network classifier. We compare optimizers that are given the same number of jobs and the same shots per job, and that differ only in how they fill each job. This comparison treats hybrid quantum-classical training as an execution-efficiency problem: how to obtain useful optimization progress from each costly interaction with a quantum processor. The resulting perspective connects optimizer design to overhead amortization and efficient use of accelerator resources. A direct port of a batched pattern search that worked for the quantum approximate optimization algorithm stalls once the model has a few dozen parameters. What does transfer is the principle behind it: fill each job with a batch of parameter vectors that is fixed before the job is sent. For circuits built from Pauli rotations the natural batch is the parameter-shift stencil. We show that its gradient error falls as the inverse of the shots per job, while the error of simultaneous-perturbation methods (SPSA) has a floor that no number of shots removes. In simulation, batched parameter-shift training needs between a fifth and a half of the jobs SPSA needs, for 24 to 48 parameters, and the ordering holds at two noise levels. A new structure-aware stencil search reads the same kind of rows as sinusoids and also beats SPSA. On IBM's ibm_fez processor, in three interleaved runs at six and eight qubits, the batched optimizer reached SPSA's final training loss within 4 to 20 of its 50 jobs. Its test accuracy was clearly higher at eight qubits; at six qubits the difference was not significant. All hardware numbers can be re-derived from the raw shots.

DOI
10.48550/arxiv.2611.16365

Related papers

An Empirical Comparison of Optimizers for Quantum Machine Learning with SPSA-Based Gradients

Marco Wiedmann, Marc Hölle, Maniraman Periyasamy +5 more

Variational quantum algorithms (VQAs) have attracted a lot of attention from the quantum computing community for the last few years. Their hybrid quantum-classical nature with relatively shallow quantum circuits makes them a promising platform for demonstrating the capabilities of noisy intermediate scale quantum (NISQ) devices. Although the classical machine learning community focuses on gradient-based parameter optimization, finding near-exact gradients for variational quantum circuits (VQCs) with the parameter-shift rule introduces a large sampling overhead. Therefore, gradient-free optimizers have gained popularity in quantum machine learning circles. Among the most promising candidates is the simultaneous perturbation stochastic approximation (SPSA) algorithm, due to its low computational cost and inherent noise resilience. We introduce a novel approach that uses the approximated gradient from SPSA in combination with state-of-the-art gradient-based classical optimizers. We demonstrate numerically that this out-performs both standard SPSA and the parameter-shift rule in terms of convergence rate and absolute error in simple regression tasks. The improvement of our novel approach over SPSA with stochastic gradient descent (SGD) is even amplified when shot- and hardware-noise are taken into account. We also demonstrate that error mitigation does not significantly affect our results.

arXiv:2305.00224•IEEE International Conference on Quantum Computing and Engineering (QCE)•Sep 17, 2023•38 citations

Hybrid Quantum-Classical Training in Quantum Neural Networks: A Framework for Bridging Parameterized Quantum Circuits and Classical Optimization

Matthew Busel

Quantum Neural Networks (QNNs) represent a compelling convergence of quantum mechanics and machine learning, yet their practical deployment remains constrained by noise, limited qubit counts, and the cost of full quantum execution. Hybrid quantum-classical (HQC) training offers a pragmatic path forward: parameterized quantum circuits (PQCs) process information in exponentially large Hilbert spaces while classical optimizers update parameters via gradient methods accessible on today's hardware. We present a modular HQC training framework, formalising the parameter-shift rule for hardware-compatible gradient estimation, characterising the barren-plateau phenomenon through a variance-decay analysis, and introducing noise-aware training strategies for NISQ devices. Empirical results from state-vector and shot-based simulation confirm measurable quantum advantage on quantum-structured tasks, with the phase classification benchmark improving +12.1% over a classical MLP baseline. The accompanying open-source library provides a clean abstraction between quantum simulation and classical optimisation, serving as a foundation for near-term quantum machine learning research.

•Mar 3, 2026•0 citations

Learning to learn with quantum neural networks via classical neural networks

Guillaume Verdon, Michael Broughton, Jarrod R. McClean +5 more

Quantum Neural Networks (QNNs) are a promising variational learning paradigm with applications to near-term quantum processors, however they still face some significant challenges. One such challenge is finding good parameter initialization heuristics that ensure rapid and consistent convergence to local minima of the parameterized quantum circuit landscape. In this work, we train classical neural networks to assist in the quantum learning process, also know as meta-learning, to rapidly find approximate optima in the parameter landscape for several classes of quantum variational algorithms. Specifically, we train classical recurrent neural networks to find approximately optimal parameters within a small number of queries of the cost function for the Quantum Approximate Optimization Algorithm (QAOA) for MaxCut, QAOA for Sherrington-Kirkpatrick Ising model, and for a Variational Quantum Eigensolver for the Hubbard model. By initializing other optimizers at parameter values suggested by the classical neural network, we demonstrate a significant improvement in the total number of optimization iterations required to reach a given accuracy. We further demonstrate that the optimization strategies learned by the neural network generalize well across a range of problem instance sizes. This opens up the possibility of training on small, classically simulatable problem instances, in order to initialize larger, classically intractably simulatable problem instances on quantum devices, thereby significantly reducing the number of required quantum-classical optimization iterations.

arXiv:1907.05415•Jul 11, 2019•274 citations

Benchmarking Modular Optimization Strategies for Parameterized Quantum Circuits

Carla Cotea, Stefan Balauca, Andreea Arusoaie

Training parameterized quantum circuits requires balancing optimization progress with limited evaluation budgets and the statistical uncertainty of near-term quantum hardware. We present a modular benchmarking framework that separates quantum search-direction estimation from classical parameter-update rules, allowing their interactions and sensitivities to be examined independently. We evaluate a suite of gradient-based, stochastic, and derivative-free optimizers across combinatorial optimization using the Quantum Approximate Optimization Algorithm (QAOA), supervised quantum machine learning using Iris classification and a quantum convolutional neural network (QCNN) for binary MNIST classification, and quantum chemistry using the variational quantum eigensolver (VQE) for molecular hydrogen. By separating objective evaluations from sample-circuit costs under finite-shot simulation and physical execution on a 156-qubit processor, we compare optimizer behavior across workloads with different parameter counts and measurement requirements. Four selected simulator and hardware case studies report terminal and best observed objectives separately, alongside post-training classifier accuracies. The comparisons include selected MaxCut and hydrogen trajectories and illustrate workload-dependent objective changes and evaluation costs, while the simulator benchmark compares both peak and mean performance across three seeds. The hardware runs do not establish an optimizer ranking or isolate a causal effect of device noise.

quant-pharXiv:2610.10254•Oct 8, 2026•0 citations

Image Classification on IBM Quantum Computers

Junghoon Justin Park, Jiook Cha, Jun-gyeong Park +2 more

Quantum machine learning on real noisy intermediate-scale quantum (NISQ) hardware has remained largely confined to binary or few-class tasks, limited by the cost of on-hardware training and the underuse of large devices at inference. We present a unified framework that classifies ten-class MNIST end-to-end on a $127$-qubit IBM Eagle processor, with three central contributions. First, a two-phase protocol decouples a gradient-based classical optimization of the encoder and readout from a gradient-free optimization of the quantum parameters, removing the parameter-shift gradient cost that makes on-hardware training impractical. Second, we introduce Quantum Multi-Programming to a trained quantum classifier for the first time, packing multiple circuit copies onto one device to deliver parallel inference at no mean-accuracy cost while cutting quantum-processor job submissions proportionally. Third, a controlled comparison shows that on-hardware fine-tuning yields no measurable accuracy gain, motivating a practical NISQ workflow: train on a classical simulator and reserve the hardware for inference only. Benchmarked against a matched-capacity classical network, the quantum module shows no per-parameter accuracy advantage at this scale; we therefore frame the work as a feasibility-and-workflow demonstration for multi-class quantum image classification on current hardware.

quant-pharXiv:2607.17705•Jul 21, 2026•0 citations

SPSA Hyperparameter Tuning for Variational Quantum Natural Language Inference

Nayan D'Souza, Christopher J. Agostino

Training variational quantum models requires choosing between parameter-shift gradients, which are exact but cost $O(P)$ forward evaluations, and simultaneous perturbation stochastic approximation (SPSA), which uses only two samples but produces high-variance estimates that can degrade optimisation on small supervised tasks. Whether the cheap gradient is usable depends on the variance that results from different choices of the SPSA perturbation scale, learning rate, and gain-decay schedule. We varied those quantities across a broad grid on a 6-qubit, 60-parameter QNLI classifier and compared the best configurations to parameter-shift AdamW and BuresQNG. AdamW-style SPSA with $c_0=0.01$, $η=0.10$, $γ=0.10$ reached $55\% \pm 11\%$ test accuracy, improving over the default configuration ($49\% \pm 6\%$) but remaining 16-19 percentage points below the parameter-shift baselines because the two-sample SPSA gradient estimate has too much variance for reliable optimisation of 60 parameters in 40 epochs. Classical-gain SPSA and Bures-preconditioned SPSA performed worse, at $51\%$ and $46\%$ respectively. Bures-preconditioning a noisy two-sample SPSA gradient amplifies perturbation noise.

quant-pharXiv:2608.16939•Aug 12, 2026•0 citations
When a quantum processor is reached through a cloud queue, a variational algorithm pays for every round trip, and a single round trip can carry many circuits. The classical optimizer therefore decides how much is learned per job, not only per shot. We study this for a quantum neural network classifier. We compare optimizers that are given the same number of jobs and the same shots per job, and that differ only in how they fill each job. This comparison treats hybrid quantum-classical training as an execution-efficiency problem: how to obtain useful optimization progress from each costly interaction with a quantum processor. The resulting perspective connects optimizer design to overhead amortization and efficient use of accelerator resources. A direct port of a batched pattern search that worked for the quantum approximate optimization algorithm stalls once the model has a few dozen parameters. What does transfer is the principle behind it: fill each job with a batch of parameter vectors that is fixed before the job is sent. For circuits built from Pauli rotations the natural batch is the parameter-shift stencil. We show that its gradient error falls as the inverse of the shots per job, while the error of simultaneous-perturbation methods (SPSA) has a floor that no number of shots removes. In simulation, batched parameter-shift training needs between a fifth and a half of the jobs SPSA needs, for 24 to 48 parameters, and the ordering holds at two noise levels. A new structure-aware stencil search reads the same kind of rows as sinusoids and also beats SPSA. On IBM's ibm_fez processor, in three interleaved runs at six and eight qubits, the batched optimizer reached SPSA's final training loss within 4 to 20 of its 50 jobs. Its test accuracy was clearly higher at eight qubits; at six qubits the difference was not significant. All hardware numbers can be re-derived from the raw shots.
Variational quantum algorithms (VQAs) have attracted a lot of attention from the quantum computing community for the last few years. Their hybrid quantum-classical nature with relatively shallow quantum circuits makes them a promising platform for demonstrating the capabilities of noisy intermediate scale quantum (NISQ) devices. Although the classical machine learning community focuses on gradient-based parameter optimization, finding near-exact gradients for variational quantum circuits (VQCs) with the parameter-shift rule introduces a large sampling overhead. Therefore, gradient-free optimizers have gained popularity in quantum machine learning circles. Among the most promising candidates is the simultaneous perturbation stochastic approximation (SPSA) algorithm, due to its low computational cost and inherent noise resilience. We introduce a novel approach that uses the approximated gradient from SPSA in combination with state-of-the-art gradient-based classical optimizers. We demonstrate numerically that this out-performs both standard SPSA and the parameter-shift rule in terms of convergence rate and absolute error in simple regression tasks. The improvement of our novel approach over SPSA with stochastic gradient descent (SGD) is even amplified when shot- and hardware-noise are taken into account. We also demonstrate that error mitigation does not significantly affect our results.
Quantum Neural Networks (QNNs) represent a compelling convergence of quantum mechanics and machine learning, yet their practical deployment remains constrained by noise, limited qubit counts, and the cost of full quantum execution. Hybrid quantum-classical (HQC) training offers a pragmatic path forward: parameterized quantum circuits (PQCs) process information in exponentially large Hilbert spaces while classical optimizers update parameters via gradient methods accessible on today's hardware. We present a modular HQC training framework, formalising the parameter-shift rule for hardware-compatible gradient estimation, characterising the barren-plateau phenomenon through a variance-decay analysis, and introducing noise-aware training strategies for NISQ devices. Empirical results from state-vector and shot-based simulation confirm measurable quantum advantage on quantum-structured tasks, with the phase classification benchmark improving +12.1% over a classical MLP baseline. The accompanying open-source library provides a clean abstraction between quantum simulation and classical optimisation, serving as a foundation for near-term quantum machine learning research.
Quantum Neural Networks (QNNs) are a promising variational learning paradigm with applications to near-term quantum processors, however they still face some significant challenges. One such challenge is finding good parameter initialization heuristics that ensure rapid and consistent convergence to local minima of the parameterized quantum circuit landscape. In this work, we train classical neural networks to assist in the quantum learning process, also know as meta-learning, to rapidly find approximate optima in the parameter landscape for several classes of quantum variational algorithms. Specifically, we train classical recurrent neural networks to find approximately optimal parameters within a small number of queries of the cost function for the Quantum Approximate Optimization Algorithm (QAOA) for MaxCut, QAOA for Sherrington-Kirkpatrick Ising model, and for a Variational Quantum Eigensolver for the Hubbard model. By initializing other optimizers at parameter values suggested by the classical neural network, we demonstrate a significant improvement in the total number of optimization iterations required to reach a given accuracy. We further demonstrate that the optimization strategies learned by the neural network generalize well across a range of problem instance sizes. This opens up the possibility of training on small, classically simulatable problem instances, in order to initialize larger, classically intractably simulatable problem instances on quantum devices, thereby significantly reducing the number of required quantum-classical optimization iterations.
Training parameterized quantum circuits requires balancing optimization progress with limited evaluation budgets and the statistical uncertainty of near-term quantum hardware. We present a modular benchmarking framework that separates quantum search-direction estimation from classical parameter-update rules, allowing their interactions and sensitivities to be examined independently. We evaluate a suite of gradient-based, stochastic, and derivative-free optimizers across combinatorial optimization using the Quantum Approximate Optimization Algorithm (QAOA), supervised quantum machine learning using Iris classification and a quantum convolutional neural network (QCNN) for binary MNIST classification, and quantum chemistry using the variational quantum eigensolver (VQE) for molecular hydrogen. By separating objective evaluations from sample-circuit costs under finite-shot simulation and physical execution on a 156-qubit processor, we compare optimizer behavior across workloads with different parameter counts and measurement requirements. Four selected simulator and hardware case studies report terminal and best observed objectives separately, alongside post-training classifier accuracies. The comparisons include selected MaxCut and hydrogen trajectories and illustrate workload-dependent objective changes and evaluation costs, while the simulator benchmark compares both peak and mean performance across three seeds. The hardware runs do not establish an optimizer ranking or isolate a causal effect of device noise.
Quantum machine learning on real noisy intermediate-scale quantum (NISQ) hardware has remained largely confined to binary or few-class tasks, limited by the cost of on-hardware training and the underuse of large devices at inference. We present a unified framework that classifies ten-class MNIST end-to-end on a 127127127-qubit IBM Eagle processor, with three central contributions. First, a two-phase protocol decouples a gradient-based classical optimization of the encoder and readout from a gradient-free optimization of the quantum parameters, removing the parameter-shift gradient cost that makes on-hardware training impractical. Second, we introduce Quantum Multi-Programming to a trained quantum classifier for the first time, packing multiple circuit copies onto one device to deliver parallel inference at no mean-accuracy cost while cutting quantum-processor job submissions proportionally. Third, a controlled comparison shows that on-hardware fine-tuning yields no measurable accuracy gain, motivating a practical NISQ workflow: train on a classical simulator and reserve the hardware for inference only. Benchmarked against a matched-capacity classical network, the quantum module shows no per-parameter accuracy advantage at this scale; we therefore frame the work as a feasibility-and-workflow demonstration for multi-class quantum image classification on current hardware.
Training variational quantum models requires choosing between parameter-shift gradients, which are exact but cost O(P)O(P)O(P) forward evaluations, and simultaneous perturbation stochastic approximation (SPSA), which uses only two samples but produces high-variance estimates that can degrade optimisation on small supervised tasks. Whether the cheap gradient is usable depends on the variance that results from different choices of the SPSA perturbation scale, learning rate, and gain-decay schedule. We varied those quantities across a broad grid on a 6-qubit, 60-parameter QNLI classifier and compared the best configurations to parameter-shift AdamW and BuresQNG. AdamW-style SPSA with c0=0.01c_0=0.01c0​=0.01, η=0.10η=0.10η=0.10, γ=0.10γ=0.10γ=0.10 reached 55%±11%55\% \pm 11\%55%±11% test accuracy, improving over the default configuration (49%±6%49\% \pm 6\%49%±6%) but remaining 16-19 percentage points below the parameter-shift baselines because the two-sample SPSA gradient estimate has too much variance for reliable optimisation of 60 parameters in 40 epochs. Classical-gain SPSA and Bures-preconditioned SPSA performed worse, at 51%51\%51% and 46%46\%46% respectively. Bures-preconditioning a noisy two-sample SPSA gradient amplifies perturbation noise.