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:2610.00841

Neural Fourier Surrogates for Data Reuploading Quantum Neural Networks

Oliver Knitter, Jonathan Mei, Sang Hyub Kim, Chi Chen, Masako Yamada, Martin Roetteler

quant-phcs.LGSubmitted Sep 30, 2026• Announced Oct 2, 2026• 1 citation
arXivHTMLPDF

Abstract

For quantum machine learning, the exact boundary between classical and quantum advantage is still poorly understood. Direct comparison between quantum neural networks (QNNs) and existing classical models, which encompass fundamentally different function classes, often fails to provide broader insight into the difference between the two. Inspired by the techniques of Neural Quantum States and Random Fourier Features, this work introduces Neural Fourier Surrogates (NFS), a stochastic classical neural network architecture for efficiently learning coefficients over the same finite Fourier series support as quantum neural networks. Testing on a selection of tabular benchmark datasets, we find that NFS is an effective classifier architecture broadly competitive with established classical baselines, including a comparable Random Fourier Features model, and possessing comparable performance to data-reuploading QNNs; combined with additional analysis comparing the learned Fourier spectra of QNNs and NFS on synthetic data, these results establish NFS as a natural classical baseline for evaluating QNN performance.

AI Summary

Data-reuploading quantum neural networks (QNNs) represent finite Fourier series, motivating classical comparisons that share their frequency support rather than use unrelated function classes. This paper introduces Neural Fourier Surrogates (NFS), which learn Fourier coefficients implicitly instead of enumerating the exponentially large frequency space. NFS combines an autoregressive transformer for the normalized coefficient magnitudes, a feedforward phase network, and a trainable coefficient $\ell_1$ norm. Frequency sampling provides Monte Carlo estimates of outputs and gradients; symmetry constraints ensure real outputs. A Gaussian-prior regularizer approximates QNNs’ preference for lower frequencies. NFS shares the frequency support of angle-encoded QNNs but permits coefficients unavailable to the fixed-observable circuits tested. Experiments cover 15 tabular classification datasets with 97–13,000 training examples, reporting means and standard deviations over five seeds. NFS is compared with six classical baselines, including random Fourier features, without demonstrating superior performance. Comparisons with noiseless simulated QNNs use six datasets at encoding depth $L=4$, plus width sweeps over 4–12 encoded dimensions on three datasets. Performance is qualitatively similar, with neither architecture consistently outperforming the other. Synthetic QNN-generated targets test spectral capacity and recovery: increasing encoding depth generally improves fitting, while correlations between learned and target frequency distributions cluster near zero or one. Stated limitations include small-scale benchmarks, higher training cost than conventional classical baselines without detailed runtime analysis, unmatched parameter counts, and differing QNN/NFS inductive biases. Reusing samples for outputs and gradients introduces finite-sample bias. Comparisons with quantum-hardware implementations remain future work.

Little sign of AI-generated textAI use disclosedIncrementalSmall numerics onlyOverclaimsMissing comparison
Generative AI (Anthropic Claude) was used by the authors for purposes of project and experiment ideation, benchmark code generation, and general manuscript outlining and proofreading.

Central claim may not hold: The claimed scalable efficiency is not established because Monte Carlo output accuracy depends on the Fourier coefficient 1-norm, while neither that norm nor gradient-estimator variance is controlled as dimension grows. The specified antisymmetric phase forces the zero-frequency coefficient to be nonnegative, contradicting the claimed full expressivity over real Fourier series and excluding negative constants that QNNs can represent. The spectral-recovery experiment does not establish that alternative spectra learned by NFS are inaccessible to the compared QNN architecture.

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

Comments
10 pages, 5 figures, 2 tables

Related papers

Benchmarking Quantum Surrogate Models on Scarce and Noisy Data

Jonas Stein, M H POPPEL, Philip Adamczyk +9 more

Surrogate models are ubiquitously used in industry and academia to efficiently approximate given black box functions. As state-of-the-art methods from classical machine learning frequently struggle to solve this problem accurately for the often scarce and noisy data sets in practical applications, investigating novel approaches is of great interest. Motivated by recent theoretical results indicating that quantum neural networks (QNNs) have the potential to outperform their classical analogs in the presence of scarce and noisy data, we benchmark their qualitative performance for this scenario empirically. Our contribution displays the first application-centered approach of using QNNs as surrogate models on higher dimensional, real world data. When compared to a classical artificial neural network with a similar number of parameters, our QNN demonstrates significantly better results for noisy and scarce data, and thus motivates future work to explore this potential quantum advantage in surrogate modelling. Finally, we demonstrate the performance of current NISQ hardware experimentally and estimate the gate fidelities necessary to replicate our simulation results.

arXiv:2306.05042•International Conference on Agents and Artificial Intelligence (ICAART)•Jan 1, 2024•7 citations

On Dequantization of Supervised Quantum Machine Learning via Random Fourier Features

Mehrad Sahebi, Alice Barthe, Yudai Suzuki +2 more

In the quest for quantum advantage, a central question is under what conditions can classical algorithms achieve a performance comparable to quantum algorithms--a concept known as dequantization. Random Fourier features (RFFs) have demonstrated potential for dequantizing certain quantum neural networks (QNNs) applied to regression tasks, but their applicability to other learning problems and architectures remained unexplored. In this work, we derive bounds on the true risk gap between classical RFF models and quantum models for regression and classification tasks with both QNN and quantum kernel architectures. Furthermore, we provide sufficient conditions under which this gap is small and thus the quantum system can be dequantized via the RFF method. We support our findings with numerical experiments that illustrate the practical dequantization of existing quantum kernel-based methods. Our findings not only broaden the applicability of RFF-dequantization but also enhance the understanding of potential quantum advantages in practical machine-learning tasks.

arXiv:2505.15902•May 21, 2025•22 citations

Parameter efficient hybrid spiking-quantum convolutional neural network with surrogate gradient and quantum data-reupload

Nhan Trong Luu, Duong Trung Luu, Nam Ngoc Pham +1 more

The rapid advancement of artificial intelligence (AI) and deep learning (DL) has catalyzed the emergence of several optimization-driven subfields, notably neuromorphic computing and quantum machine learning. Leveraging the differentiable nature of hybrid models, researchers have explored their potential to address complex problems through unified optimization strategies. One such development is the Spiking Quantum Neural Network (SQNN), which combines principles from spiking neural networks (SNNs) and quantum computing. However, existing SQNN implementations often depend on pretrained SNNs due to the non-differentiable nature of spiking activity and the limited scalability of current SNN encoders. In this work, we propose a novel architecture—Spiking-Quantum Data Re-upload Convolutional Neural Network (SQDR-CNN)—that enables joint training of convolutional SNNs and quantum circuits within a single backpropagation framework. Unlike its predecessor, SQDR-CNN allow convergence to reasonable performance without the reliance of pretrained spiking encoder and subsetting datasets. We also clarified some theoretical foundations, testing new design using quantum data-reupload with different training algorithm-initialization and evaluate the performance of the proposed model under noisy simulated quantum environments. As a result, we were able to achieve 86% of the mean top-performing accuracy of the state of the art (SOTA) SNN baselines, yet uses only 0.5% of the smallest spiking model’s parameters. Through this integration of neuromorphic and quantum paradigms, we aim to open new research directions and foster technological progress in multi-modal, learnable systems.

arXiv:2512.03895•PeerJ Computer Science•Feb 24, 2026•10 citations

A Comprehensive Analysis of Accuracy and Robustness in Quantum Neural Networks

Ban Q. Tran, Duong M. Chu, Hai T. D. Pham +3 more

Quantum Machine Learning (QML) has recently emerged as a highly promising research frontier. Within this domain, Quantum Neural Networks (QNNs),characterized by Variational Quantum Circuits (VQCs) at their core and featuring layers of quantum gates optimized by classical algorithms, have garnered significant attention. However, a rigorous and exhaustive evaluation of their practical performance remains largely incomplete. In this study, we conduct a comprehensive comparative analysis of three prominent hybrid classical-quantum architectures: Quantum Convolutional Neural Networks (QCNN), Quantum Recurrent Neural Networks (QRNN), and Quantum Vision Transformers (QViT), focusing on the critical dimensions of generalization, accuracy, and robustness. Our findings provide novel insights that address previous evaluative gaps. Notably, while these models exhibit exceptional performance on low-feature datasets such as MNIST, their learning efficacy degrades significantly when transitioned to high-feature datasets. Furthermore, convolutional-based models like QCNN appear less effective on high-dimensional data than other machine learning architectures. Additionally, while all models are susceptible to adversarial noise, traditional architectures, such as recurrent and convolutional networks, demonstrate superior resilience. Conversely, in the presence of quantum noise, the transformer-based architecture proves its strength by maintaining high robustness against measurement noise, channel noise, and finite-shot effects, whereas other architectures suffer marked performance declines. These results provide a granular perspective on the current state of the field and underscore the critical importance of tailoring model selection to the constraints of contemporary Noisy Intermediate-Scale Quantum (NISQ) environments.

quant-pharXiv:2604.26110•Apr 30, 2026•1 citation

Neural quantum kernels: Training quantum kernels with quantum neural networks

Pablo Rodriguez-Grasa, Yue Ban, Mikel Sanz

Quantum and classical machine learning have been naturally connected through kernel methods, which have also served as proof-of-concept for quantum advantage. Quantum embeddings encode classical data into quantum feature states, enabling the construction of embedding quantum kernels (EQKs) by measuring vector similarities and projected quantum kernels (PQKs) through projections of these states. However, in both approaches, the model is influenced by the choice of the embedding. In this work, we propose using the training of a quantum neural network (QNN) to construct neural quantum kernels, specifically neural EQKs and neural PQKs—problem-inspired kernel functions. Unlike previous approaches, our method requires the kernel matrix to be constructed only once, significantly reducing computational overhead. To achieve this, we introduce a scalable training method for an n -qubit data reuploading QNN. Furthermore, we demonstrate neural quantum kernels can alleviate exponential concentration and enhance generalization capabilities compared to problem-agnostic kernels, positioning them as a scalable and robust solution for quantum machine learning applications.

arXiv:2401.04642•Physical Review Research•May 20, 2025•30 citations

A Comparative Effectiveness Study of Quantum Machine Learning Algorithms

Caleb Dow, Daniel Tebor, Gayathri Priya Kolavennu +1 more

This research provides a rigorous empirical benchmark comparing classical and quantum machine learning algorithms across PCA, Random Forest, Neural Networks, and CNNs. We evaluate performance on four datasets, revealing quantum advantages only in specific scenarios: quantum neural networks achieved 34-46% better precision/recall on imbalanced classification, while classical methods dominated in efficiency and scalability. Critical findings include quantum algorithms requiring 57,500x more training time and becoming intractable beyond 8-10 qubits. Our implementation methodology and performance baselines provide practitioners with concrete guidance for algorithm selection in the current NISQ era.

IEEE International Conference on Computing (ICOCO)•Oct 6, 2025•0 citations
For quantum machine learning, the exact boundary between classical and quantum advantage is still poorly understood. Direct comparison between quantum neural networks (QNNs) and existing classical models, which encompass fundamentally different function classes, often fails to provide broader insight into the difference between the two. Inspired by the techniques of Neural Quantum States and Random Fourier Features, this work introduces Neural Fourier Surrogates (NFS), a stochastic classical neural network architecture for efficiently learning coefficients over the same finite Fourier series support as quantum neural networks. Testing on a selection of tabular benchmark datasets, we find that NFS is an effective classifier architecture broadly competitive with established classical baselines, including a comparable Random Fourier Features model, and possessing comparable performance to data-reuploading QNNs; combined with additional analysis comparing the learned Fourier spectra of QNNs and NFS on synthetic data, these results establish NFS as a natural classical baseline for evaluating QNN performance.
Data-reuploading quantum neural networks (QNNs) represent finite Fourier series, motivating classical comparisons that share their frequency support rather than use unrelated function classes. This paper introduces Neural Fourier Surrogates (NFS), which learn Fourier coefficients implicitly instead of enumerating the exponentially large frequency space.
NFS combines an autoregressive transformer for the normalized coefficient magnitudes, a feedforward phase network, and a trainable coefficient ℓ1\ell_1ℓ1​ norm. Frequency sampling provides Monte Carlo estimates of outputs and gradients; symmetry constraints ensure real outputs. A Gaussian-prior regularizer approximates QNNs’ preference for lower frequencies. NFS shares the frequency support of angle-encoded QNNs but permits coefficients unavailable to the fixed-observable circuits tested.
Experiments cover 15 tabular classification datasets with 97–13,000 training examples, reporting means and standard deviations over five seeds. NFS is compared with six classical baselines, including random Fourier features, without demonstrating superior performance. Comparisons with noiseless simulated QNNs use six datasets at encoding depth L=4L=4L=4, plus width sweeps over 4–12 encoded dimensions on three datasets. Performance is qualitatively similar, with neither architecture consistently outperforming the other. Synthetic QNN-generated targets test spectral capacity and recovery: increasing encoding depth generally improves fitting, while correlations between learned and target frequency distributions cluster near zero or one.
Stated limitations include small-scale benchmarks, higher training cost than conventional classical baselines without detailed runtime analysis, unmatched parameter counts, and differing QNN/NFS inductive biases. Reusing samples for outputs and gradients introduces finite-sample bias. Comparisons with quantum-hardware implementations remain future work.
Surrogate models are ubiquitously used in industry and academia to efficiently approximate given black box functions. As state-of-the-art methods from classical machine learning frequently struggle to solve this problem accurately for the often scarce and noisy data sets in practical applications, investigating novel approaches is of great interest. Motivated by recent theoretical results indicating that quantum neural networks (QNNs) have the potential to outperform their classical analogs in the presence of scarce and noisy data, we benchmark their qualitative performance for this scenario empirically. Our contribution displays the first application-centered approach of using QNNs as surrogate models on higher dimensional, real world data. When compared to a classical artificial neural network with a similar number of parameters, our QNN demonstrates significantly better results for noisy and scarce data, and thus motivates future work to explore this potential quantum advantage in surrogate modelling. Finally, we demonstrate the performance of current NISQ hardware experimentally and estimate the gate fidelities necessary to replicate our simulation results.
In the quest for quantum advantage, a central question is under what conditions can classical algorithms achieve a performance comparable to quantum algorithms--a concept known as dequantization. Random Fourier features (RFFs) have demonstrated potential for dequantizing certain quantum neural networks (QNNs) applied to regression tasks, but their applicability to other learning problems and architectures remained unexplored. In this work, we derive bounds on the true risk gap between classical RFF models and quantum models for regression and classification tasks with both QNN and quantum kernel architectures. Furthermore, we provide sufficient conditions under which this gap is small and thus the quantum system can be dequantized via the RFF method. We support our findings with numerical experiments that illustrate the practical dequantization of existing quantum kernel-based methods. Our findings not only broaden the applicability of RFF-dequantization but also enhance the understanding of potential quantum advantages in practical machine-learning tasks.
The rapid advancement of artificial intelligence (AI) and deep learning (DL) has catalyzed the emergence of several optimization-driven subfields, notably neuromorphic computing and quantum machine learning. Leveraging the differentiable nature of hybrid models, researchers have explored their potential to address complex problems through unified optimization strategies. One such development is the Spiking Quantum Neural Network (SQNN), which combines principles from spiking neural networks (SNNs) and quantum computing. However, existing SQNN implementations often depend on pretrained SNNs due to the non-differentiable nature of spiking activity and the limited scalability of current SNN encoders. In this work, we propose a novel architecture—Spiking-Quantum Data Re-upload Convolutional Neural Network (SQDR-CNN)—that enables joint training of convolutional SNNs and quantum circuits within a single backpropagation framework. Unlike its predecessor, SQDR-CNN allow convergence to reasonable performance without the reliance of pretrained spiking encoder and subsetting datasets. We also clarified some theoretical foundations, testing new design using quantum data-reupload with different training algorithm-initialization and evaluate the performance of the proposed model under noisy simulated quantum environments. As a result, we were able to achieve 86% of the mean top-performing accuracy of the state of the art (SOTA) SNN baselines, yet uses only 0.5% of the smallest spiking model’s parameters. Through this integration of neuromorphic and quantum paradigms, we aim to open new research directions and foster technological progress in multi-modal, learnable systems.
Quantum Machine Learning (QML) has recently emerged as a highly promising research frontier. Within this domain, Quantum Neural Networks (QNNs),characterized by Variational Quantum Circuits (VQCs) at their core and featuring layers of quantum gates optimized by classical algorithms, have garnered significant attention. However, a rigorous and exhaustive evaluation of their practical performance remains largely incomplete. In this study, we conduct a comprehensive comparative analysis of three prominent hybrid classical-quantum architectures: Quantum Convolutional Neural Networks (QCNN), Quantum Recurrent Neural Networks (QRNN), and Quantum Vision Transformers (QViT), focusing on the critical dimensions of generalization, accuracy, and robustness. Our findings provide novel insights that address previous evaluative gaps. Notably, while these models exhibit exceptional performance on low-feature datasets such as MNIST, their learning efficacy degrades significantly when transitioned to high-feature datasets. Furthermore, convolutional-based models like QCNN appear less effective on high-dimensional data than other machine learning architectures. Additionally, while all models are susceptible to adversarial noise, traditional architectures, such as recurrent and convolutional networks, demonstrate superior resilience. Conversely, in the presence of quantum noise, the transformer-based architecture proves its strength by maintaining high robustness against measurement noise, channel noise, and finite-shot effects, whereas other architectures suffer marked performance declines. These results provide a granular perspective on the current state of the field and underscore the critical importance of tailoring model selection to the constraints of contemporary Noisy Intermediate-Scale Quantum (NISQ) environments.
Quantum and classical machine learning have been naturally connected through kernel methods, which have also served as proof-of-concept for quantum advantage. Quantum embeddings encode classical data into quantum feature states, enabling the construction of embedding quantum kernels (EQKs) by measuring vector similarities and projected quantum kernels (PQKs) through projections of these states. However, in both approaches, the model is influenced by the choice of the embedding. In this work, we propose using the training of a quantum neural network (QNN) to construct neural quantum kernels, specifically neural EQKs and neural PQKs—problem-inspired kernel functions. Unlike previous approaches, our method requires the kernel matrix to be constructed only once, significantly reducing computational overhead. To achieve this, we introduce a scalable training method for an n -qubit data reuploading QNN. Furthermore, we demonstrate neural quantum kernels can alleviate exponential concentration and enhance generalization capabilities compared to problem-agnostic kernels, positioning them as a scalable and robust solution for quantum machine learning applications.
This research provides a rigorous empirical benchmark comparing classical and quantum machine learning algorithms across PCA, Random Forest, Neural Networks, and CNNs. We evaluate performance on four datasets, revealing quantum advantages only in specific scenarios: quantum neural networks achieved 34-46% better precision/recall on imbalanced classification, while classical methods dominated in efficiency and scalability. Critical findings include quantum algorithms requiring 57,500x more training time and becoming intractable beyond 8-10 qubits. Our implementation methodology and performance baselines provide practitioners with concrete guidance for algorithm selection in the current NISQ era.