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:2610.39577 · IEEE International Conference on Quantum Artificial Intelligence (QAI)

Adiabatic training of VQAs for entanglement detection

E. Acosta, C. Cano, D. Manzano, G. Botella

quant-phSubmitted Oct 29, 2026• Announced Oct 30, 2026• 1 citation
arXivHTMLPDF

Abstract

This paper introduces a hybrid quantum framework for training Variational Quantum Algorithms (VQAs) aimed at addressing the fundamental problem of entanglement detection using adiabatic quantum computing. This problem is of foundational significance in quantum information science and has been approached through both classical and quantum methodologies. We propose an adiabatic-inspired optimization strategy for training VQAs, which combines a discretization of the parameter space with a hierarchical segmentation scheme. The optimization problem is formulated as a Quadratic Unconstrained Binary Optimization (QUBO) model to leverage the capabilities of quantum annealing. Preliminary results are compared against classical neural networks and conventional VQA training techniques. The findings aim to stimulate further investigation and discussion within the scientific community regarding the efficacy and applicability of the proposed approach.

AI Summary

The paper studies whether quantum-annealing-based optimization can train variational quantum circuits to classify quantum states as separable or entangled. It applies a previously introduced training framework that discretizes circuit rotation angles, partitions parameter space, and hierarchically refines the search around promising parameter choices. A state-vector mean-squared-error objective is presented as a QUBO for annealing-based optimization. The numerical study uses generated three-qubit pure states, with balanced separable and entangled labels and equal representation of bipartite-entangled, W-type, and GHZ-type examples within the entangled class. Comparisons include SPSA-trained variational circuits and a small feedforward neural network. The annealing-based configuration uses four refinement levels, three partitions per angle, and initially three discretization points per angle; the experimental settings specify 25 executions per model. Reported average training/validation accuracies are approximately $67\%/61\%$ for annealing-based training, $62.5\%/60\%$ for conventional VQA training, and $75\%/65\%$ for the neural network. Reported subclass accuracies are $65\%$ for bipartite entanglement, $60\%$ for GHZ states, and $55\%$ for W states. Training accuracy increases across successive refinement levels. The paper describes these results as preliminary. It acknowledges discretization-induced precision loss and the possibility of missing global minima. Execution-time comparisons remain unevaluated, and an evolutionary-training baseline was not completed because of computational demands. Larger-scale testing and further refinement of the search strategy are left for future work.

Some signs of AI-generated textOverclaimsSmall numerics onlyUnclear noveltyPossible error

Central claim may not hold: The proposed QUBO is not established by the displayed equations, which omit an explicit binary encoding, selection penalties, and any necessary reduction of higher-order circuit-dependent terms. The displayed training objective compares each input state with its image under Q rather than with an entanglement label, admits zero residual for the identity transformation if available, and uses ordinary squares of potentially complex residuals without defining a real-valued encoding. The claimed computational savings from simultaneous annealing evaluation lack a runtime and success-probability analysis, while the experiments do not identify the annealing device or solver and its execution settings. The experimental description misidentifies three-qubit dimensions, and its local-unitary Bell-pair generator excludes partially entangled bipartite pure states, undermining the claimed generalization to all three-qubit states.

Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.

Journal reference
Proceedings - 2025 IEEE International Conference on Quantum Artificial Intelligence, QAI 2025, 27-33
DOI
10.1109/QAI63978.2025.00012
Comments
7 pages, 8 figures

Related papers

Scalable Optimization of Binary Neural Networks Using Variational Quantum Algorithms

Anton Naim Ibrahim

I present a novel approach to training Binary Neural Networks (BiNNs) by leveraging Variational Quantum Algorithms (VQAs). My method employs the hybrid quantum-classical framework of VQAs to train simulated quantum circuits as hypernetworks that generate effective BiNN configurations. By developing approximate methods for calculating the VQA objective and gradient, I significantly scale quantum hypernetworks—over two orders of magnitude larger than previous capabilities, overcoming the constraints of exact computations. Quantum circuits are trained to produce effective BiNNs for binary classification tasks on resized and native resolution MNIST datasets. My empirical investigation extends to various VQA hyperparameters, including sample and batch sizes, circuit depth, and Multi-Basis Encoding (MBE). The results reveal that my optimized circuits consistently generate BiNN configurations with competitive cross-entropy losses. This work showcases the practical scalability of quantum hypernetworks and demonstrates significant interdisciplinary contribution in quantum machine learning, inviting further exploration into hybrid quantum-classical computational techniques.

•Aug 1, 2024•0 citations

Quantum–inspired Methods for Training Machine Learning Models

Nilesh T. Fonseka, Anuradha Mahasinghe

Machine learning model training, which ultimately optimizes a model’s cost function is usually a time- consuming and computationally intensive process on classical computers. This has been more intense due to the in- creased demand for large-scale data analysis, requiring unconventional computing paradigms like quantum computing to enhance training efficiency. Adiabatic quantum computers have excelled at solving optimization problems, which require the quadratic unconstrained binary optimization (QUBO) format of the problem of interest. In this study, the squared error minimization in the multiple linear regression model is reformulated as a QUBO problem enabling it to be solved using D-wave adiabatic quantum computers. Same formulation was used to obtain a solution using gate-based algorithms such as quantum approximate optimization algorithm (QAOA) and sampling variational quantum eigensolver (VQE) im- plemented via IBM Qiskit. The results obtained through these approaches in the context of runtime and mean squared error(MSE) were analyzed and compared to the classical approaches. Our experimental results indicate a runtime ad- vantage in the D-wave annealing approach over the classical Scikit learn regression approach. The time advantage can be observed when N>524288 compared to Sklearn Linear Regression and when N>65536 compared to Sklearn SGDRegressor. Support vector machine induced neural networks, where the margin-based entropy loss is converted into a QUBO with Lagrangian approach is also focused in this study concerning the applicability for nonlinear models.

International Journal of Information Technology and Computer Science•Dec 2, 2025•0 citations

Quantum-Inspired Optimization of Neural Network Training Dynamics

Jincheng Zhang

Training deep neural networks remains a computationally intensive and often challenging process. Traditional optimization methods, such as stochastic gradient descent (SGD) and its variants, frequently struggle to escape local minima and converge to optimal solutions, leading to suboptimal network performance. This paper proposes a novel approach to neural network training that leverages principles inspired by quantum mechanics. Specifically, we investigate the application of quantum-inspired optimization techniques, including simulated annealing and quantum walks, to dynamically adjust the training process. The core idea is to mimic the probabilistic exploration inherent in quantum systems to effectively navigate the complex loss landscape of neural networks. We formulate the training process as a quantum-inspired optimization problem, where the network's weights are treated as quantum bits (Qubits) and the optimization algorithm seeks to find the lowest energy state, analogous to finding the minimum loss. The proposed method demonstrates the potential for accelerating convergence and improving the overall accuracy of neural networks compared to conventional optimization strategies. Our analysis focuses on the key aspects of the approach, including the mathematical formulation of the quantum-inspired optimization problem and the selection of appropriate parameters to govern the exploration and exploitation phases. We present a theoretical framework for understanding the dynamics of this approach and discuss potential avenues for further research, including hybrid quantum-classical algorithms and the integration of quantum computing hardware.

•Sep 5, 2026•0 citations

A framework to evaluate the performance of Variational Quantum Algorithms

Ernesto Mamedaliev, Vladyslav Libov, Albert Nieto-Morales +2 more

Variational Quantum Algorithms (VQAs) are promising methods for solving combinatorial optimization problems on noisy intermediate-scale quantum (NISQ) devices. However, benchmarking VQAs is difficult due to their stochastic behavior and the lack of standardized performance criteria. This work introduces a general framework for evaluating VQAs applied to Quadratic Unconstrained Binary Optimization (QUBO) problems. The framework uses three complementary metrics: feasibility, quality, and reproducibility. It also introduces a quality diagram that visualizes trade-offs between success probability and computational resources. Reproducibility is formalized using Shannon entropy, and a decision rule is defined for selecting algorithms under resource constraints. As a demonstration, the framework is applied to several VQAs using Conditional Value at Risk (CVaR) cost functions and different shot counts on a 16-qubit QUBO instance. The results show how the framework supports systematic benchmarking and provides a foundation for adaptive algorithm selection in hybrid quantum-classical workflows.

quant-pharXiv:2601.18812•Jan 22, 2026•0 citations

Neural-network-encoded variational quantum algorithms

Jiaqi Miao, Chang‐Yu Hsieh, Shi‐Xin Zhang

We introduce a general framework called neural-network- (NN) encoded variational quantum algorithms (VQAs), or NNVQA for short, to address the challenges of implementing VQAs on noisy intermediate-scale quantum (NISQ) computers. Specifically, NNVQA feeds input (such as parameters of a Hamiltonian) from a given problem to a neural network and uses its outputs to parameterize an ansatz circuit for the standard VQA. Combining the strengths of NN and parameterized quantum circuits, NNVQA can accelerate the training process of VQAs and handle a broad family of related problems with varying input parameters with the pretrained NN. To concretely illustrate the merits of NNVQA, we present results on a NN variational quantum eigensolver (VQE) for solving the ground state of parameterized $XXZ$ spin models in one and two dimensions. Our results demonstrate that NNVQE is able to estimate the ground-state energies of parameterized Hamiltonians with high precision without fine tuning, and significantly reduce the overall training cost to estimate ground-state properties across the phases of the given Hamiltonian. We also employ an active learning strategy to further increase the training efficiency while maintaining prediction accuracy. These encouraging results demonstrate that NNVQAs offer an alternative hybrid quantum-classical paradigm to utilize NISQ resources for solving more realistic and challenging computational problems.

arXiv:2308.01068•Physical Review Applied•Jan 26, 2024•25 citations

Quantum-Enhanced Cloud Intelligence: Hybrid Variational Models for Next-Generation Scalable Machine Learning

Quentil Varano, Iselphine Draxmont, Zalmera O. Nyvrenn

This research investigates a quantum enhanced cloud intelligence framework based on hybrid variational models that integrate Variational Quantum Algorithms (VQAs), Parameterized Quantum Circuits (PQCs), and classical machine learning optimization. The study aims to address the scalability and computational limitations of conventional cloud-based machine learning by leveraging the expressive power of quantum feature spaces and entanglement-driven representations. A structured methodology is presented, encompassing hybrid model design, dataset preparation, quantum circuit construction, and the implementation of a cloud-integrated training loop. Performance benchmarking across high-dimensional datasets demonstrates that the proposed hybrid approach can achieve faster training, improved model accuracy, and enhanced energy efficiency compared to classical baselines. The research further outlines the practical challenges posed by NISQ-era hardware, noise sensitivity, cloud latency, and hybrid optimization instability. Despite these limitations, the findings reveal strong potential for deploying quantum-assisted intelligence in real-time analytics, complex optimization problems, healthcare diagnostics, autonomous systems, and cybersecurity applications. This study contributes a unified integration framework, novel empirical benchmarks, and a practical roadmap for advancing quantum cloud synergy, positioning hybrid variational systems as a promising foundation for next-generation scalable machine learning.

Jurnal Teknik Informatika C I T Medicom•Jul 30, 2025•0 citations
This paper introduces a hybrid quantum framework for training Variational Quantum Algorithms (VQAs) aimed at addressing the fundamental problem of entanglement detection using adiabatic quantum computing. This problem is of foundational significance in quantum information science and has been approached through both classical and quantum methodologies. We propose an adiabatic-inspired optimization strategy for training VQAs, which combines a discretization of the parameter space with a hierarchical segmentation scheme. The optimization problem is formulated as a Quadratic Unconstrained Binary Optimization (QUBO) model to leverage the capabilities of quantum annealing. Preliminary results are compared against classical neural networks and conventional VQA training techniques. The findings aim to stimulate further investigation and discussion within the scientific community regarding the efficacy and applicability of the proposed approach.
The paper studies whether quantum-annealing-based optimization can train variational quantum circuits to classify quantum states as separable or entangled. It applies a previously introduced training framework that discretizes circuit rotation angles, partitions parameter space, and hierarchically refines the search around promising parameter choices. A state-vector mean-squared-error objective is presented as a QUBO for annealing-based optimization.
The numerical study uses generated three-qubit pure states, with balanced separable and entangled labels and equal representation of bipartite-entangled, W-type, and GHZ-type examples within the entangled class. Comparisons include SPSA-trained variational circuits and a small feedforward neural network. The annealing-based configuration uses four refinement levels, three partitions per angle, and initially three discretization points per angle; the experimental settings specify 25 executions per model.
Reported average training/validation accuracies are approximately 67%/61%67\%/61\%67%/61% for annealing-based training, 62.5%/60%62.5\%/60\%62.5%/60% for conventional VQA training, and 75%/65%75\%/65\%75%/65% for the neural network. Reported subclass accuracies are 65%65\%65% for bipartite entanglement, 60%60\%60% for GHZ states, and 55%55\%55% for W states. Training accuracy increases across successive refinement levels.
The paper describes these results as preliminary. It acknowledges discretization-induced precision loss and the possibility of missing global minima. Execution-time comparisons remain unevaluated, and an evolutionary-training baseline was not completed because of computational demands. Larger-scale testing and further refinement of the search strategy are left for future work.
I present a novel approach to training Binary Neural Networks (BiNNs) by leveraging Variational Quantum Algorithms (VQAs). My method employs the hybrid quantum-classical framework of VQAs to train simulated quantum circuits as hypernetworks that generate effective BiNN configurations. By developing approximate methods for calculating the VQA objective and gradient, I significantly scale quantum hypernetworks—over two orders of magnitude larger than previous capabilities, overcoming the constraints of exact computations. Quantum circuits are trained to produce effective BiNNs for binary classification tasks on resized and native resolution MNIST datasets. My empirical investigation extends to various VQA hyperparameters, including sample and batch sizes, circuit depth, and Multi-Basis Encoding (MBE). The results reveal that my optimized circuits consistently generate BiNN configurations with competitive cross-entropy losses. This work showcases the practical scalability of quantum hypernetworks and demonstrates significant interdisciplinary contribution in quantum machine learning, inviting further exploration into hybrid quantum-classical computational techniques.
Machine learning model training, which ultimately optimizes a model’s cost function is usually a time- consuming and computationally intensive process on classical computers. This has been more intense due to the in- creased demand for large-scale data analysis, requiring unconventional computing paradigms like quantum computing to enhance training efficiency. Adiabatic quantum computers have excelled at solving optimization problems, which require the quadratic unconstrained binary optimization (QUBO) format of the problem of interest. In this study, the squared error minimization in the multiple linear regression model is reformulated as a QUBO problem enabling it to be solved using D-wave adiabatic quantum computers. Same formulation was used to obtain a solution using gate-based algorithms such as quantum approximate optimization algorithm (QAOA) and sampling variational quantum eigensolver (VQE) im- plemented via IBM Qiskit. The results obtained through these approaches in the context of runtime and mean squared error(MSE) were analyzed and compared to the classical approaches. Our experimental results indicate a runtime ad- vantage in the D-wave annealing approach over the classical Scikit learn regression approach. The time advantage can be observed when N>524288 compared to Sklearn Linear Regression and when N>65536 compared to Sklearn SGDRegressor. Support vector machine induced neural networks, where the margin-based entropy loss is converted into a QUBO with Lagrangian approach is also focused in this study concerning the applicability for nonlinear models.
Training deep neural networks remains a computationally intensive and often challenging process. Traditional optimization methods, such as stochastic gradient descent (SGD) and its variants, frequently struggle to escape local minima and converge to optimal solutions, leading to suboptimal network performance. This paper proposes a novel approach to neural network training that leverages principles inspired by quantum mechanics. Specifically, we investigate the application of quantum-inspired optimization techniques, including simulated annealing and quantum walks, to dynamically adjust the training process. The core idea is to mimic the probabilistic exploration inherent in quantum systems to effectively navigate the complex loss landscape of neural networks. We formulate the training process as a quantum-inspired optimization problem, where the network's weights are treated as quantum bits (Qubits) and the optimization algorithm seeks to find the lowest energy state, analogous to finding the minimum loss. The proposed method demonstrates the potential for accelerating convergence and improving the overall accuracy of neural networks compared to conventional optimization strategies. Our analysis focuses on the key aspects of the approach, including the mathematical formulation of the quantum-inspired optimization problem and the selection of appropriate parameters to govern the exploration and exploitation phases. We present a theoretical framework for understanding the dynamics of this approach and discuss potential avenues for further research, including hybrid quantum-classical algorithms and the integration of quantum computing hardware.
Variational Quantum Algorithms (VQAs) are promising methods for solving combinatorial optimization problems on noisy intermediate-scale quantum (NISQ) devices. However, benchmarking VQAs is difficult due to their stochastic behavior and the lack of standardized performance criteria. This work introduces a general framework for evaluating VQAs applied to Quadratic Unconstrained Binary Optimization (QUBO) problems. The framework uses three complementary metrics: feasibility, quality, and reproducibility. It also introduces a quality diagram that visualizes trade-offs between success probability and computational resources. Reproducibility is formalized using Shannon entropy, and a decision rule is defined for selecting algorithms under resource constraints. As a demonstration, the framework is applied to several VQAs using Conditional Value at Risk (CVaR) cost functions and different shot counts on a 16-qubit QUBO instance. The results show how the framework supports systematic benchmarking and provides a foundation for adaptive algorithm selection in hybrid quantum-classical workflows.
We introduce a general framework called neural-network- (NN) encoded variational quantum algorithms (VQAs), or NNVQA for short, to address the challenges of implementing VQAs on noisy intermediate-scale quantum (NISQ) computers. Specifically, NNVQA feeds input (such as parameters of a Hamiltonian) from a given problem to a neural network and uses its outputs to parameterize an ansatz circuit for the standard VQA. Combining the strengths of NN and parameterized quantum circuits, NNVQA can accelerate the training process of VQAs and handle a broad family of related problems with varying input parameters with the pretrained NN. To concretely illustrate the merits of NNVQA, we present results on a NN variational quantum eigensolver (VQE) for solving the ground state of parameterized XXZXXZXXZ spin models in one and two dimensions. Our results demonstrate that NNVQE is able to estimate the ground-state energies of parameterized Hamiltonians with high precision without fine tuning, and significantly reduce the overall training cost to estimate ground-state properties across the phases of the given Hamiltonian. We also employ an active learning strategy to further increase the training efficiency while maintaining prediction accuracy. These encouraging results demonstrate that NNVQAs offer an alternative hybrid quantum-classical paradigm to utilize NISQ resources for solving more realistic and challenging computational problems.
This research investigates a quantum enhanced cloud intelligence framework based on hybrid variational models that integrate Variational Quantum Algorithms (VQAs), Parameterized Quantum Circuits (PQCs), and classical machine learning optimization. The study aims to address the scalability and computational limitations of conventional cloud-based machine learning by leveraging the expressive power of quantum feature spaces and entanglement-driven representations. A structured methodology is presented, encompassing hybrid model design, dataset preparation, quantum circuit construction, and the implementation of a cloud-integrated training loop. Performance benchmarking across high-dimensional datasets demonstrates that the proposed hybrid approach can achieve faster training, improved model accuracy, and enhanced energy efficiency compared to classical baselines. The research further outlines the practical challenges posed by NISQ-era hardware, noise sensitivity, cloud latency, and hybrid optimization instability. Despite these limitations, the findings reveal strong potential for deploying quantum-assisted intelligence in real-time analytics, complex optimization problems, healthcare diagnostics, autonomous systems, and cybersecurity applications. This study contributes a unified integration framework, novel empirical benchmarks, and a practical roadmap for advancing quantum cloud synergy, positioning hybrid variational systems as a promising foundation for next-generation scalable machine learning.