pith. sign in

arxiv: 2510.07195 · v2 · pith:NUP4QOQSnew · submitted 2025-10-08 · 🪐 quant-ph · cs.LG

Accelerating Inference for Multilayer Neural Networks with Quantum Computers

Arthur G. Rattew , Po-Wei Huang , Naixu Guo , Lirand\"e Pira , Patrick Rebentrost This is my paper

Pith reviewed 2026-05-21 20:54 UTC · model grok-4.3

classification 🪐 quant-ph cs.LG
keywords quantum machine learningneural network inferenceresidual networksquantum algorithmsfault-tolerant quantum computingquantum random access memory
0
0 comments X

The pith

With quantum access to inputs and weights, multilayer neural networks achieve polylogarithmic inference cost.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper constructs explicit quantum circuits that implement full multilayer networks modeled on ResNet, including 2D convolutions, sigmoid activations, skip connections, and layer normalizations, all while preserving coherence. It measures inference cost in three access regimes: no special assumptions yield only quadratic speedup for shallow bilinear networks; efficient quantum access to weights alone produces quartic speedup; and efficient quantum access to both inputs and weights reduces the cost for an N-dimensional input, k residual blocks, and final pooling layer to O(polylog(N/ε)^k) operations at error ε. A sympathetic reader cares because this supplies a concrete route for future fault-tolerant quantum processors to accelerate deep-learning workloads without forcing architectural redesign.

Core claim

A network with an N-dimensional vectorized input, k residual block layers, and a final residual-linear-pooling layer can be implemented with an error of ε with O(polylog(N/ε)^k) inference cost when efficient quantum access to both the inputs and the network weights is available.

What carries the argument

Coherent quantum circuits for residual blocks that embed multi-filter 2D convolutions, sigmoid activations, skip connections, and layer normalizations, realized via quantum oracles for data and weight access.

Load-bearing premise

Efficient quantum oracles or quantum random access memory must exist that supply coherent, low-overhead access to the input data and all network weights.

What would settle it

A controlled simulation or small-scale quantum run that measures whether the total gate count for a fixed-depth ResNet-style network grows as polylog(N) rather than any polynomial in N when the input dimension N is increased while keeping error below a fixed ε.

Figures

Figures reproduced from arXiv: 2510.07195 by Arthur G. Rattew, Lirand\"e Pira, Naixu Guo, Patrick Rebentrost, Po-Wei Huang.

Figure 1
Figure 1. Figure 1: Architecture for Convolutional Neural Networks. This figure shows the architectures we consider with provable quantum complexity guarantees for inference under three regimes of quantum data access assumptions. (a) Depicts the architecture where both the inputs and network weights are provided in an efficient quantum data structure. (b) Only the network weights are provided in an efficient quantum data stru… view at source ↗
Figure 2
Figure 2. Figure 2: Generic Residual Architectural Block. This diagram illustrates the structure of a typical residual block used in deep neural networks. The input vector x is transformed through a sequence of operations: a learnable linear transformation W, a non-linear activation function f, and a residual (skip) connection that adds the original input to the transformed signal. The output is then passed through a normaliz… view at source ↗
Figure 3
Figure 3. Figure 3: Circuit for addition of VE encoded vectors. Given two unitary matrices, Uψ which is a (α, a, ϵ0)-VE for the n-qubit state |ψ⟩, and Uϕ which is a (β, b, ϵ1)-VE for the n-qubit state |ϕ⟩, define c := max(a, b). We define U˜ ψ by appropriately tensoring Uψ with Ic−a and we define U˜ ϕ by appropriately tensoring Uϕ with Ic−b, such that U˜ ψ and U˜ ϕ both act on n + c qubits. Then, the given circuit yields a VE… view at source ↗
Figure 4
Figure 4. Figure 4: Full-rank linear-pooling output block [PITH_FULL_IMAGE:figures/full_fig_p037_4.png] view at source ↗
read the original abstract

Fault-tolerant Quantum Processing Units (QPUs) promise to deliver exponential speed-ups in select computational tasks, yet their integration into modern deep learning pipelines remains unclear. In this work, we take a step towards bridging this gap by presenting the first fully-coherent quantum implementation of a multilayer neural network with non-linear activation functions. Our constructions mirror widely used deep learning architectures based on ResNet, and consist of residual blocks with multi-filter 2D convolutions, sigmoid activations, skip-connections, and layer normalizations. We analyse the complexity of inference for networks under three quantum data access regimes. Without any assumptions, we establish a quadratic speedup over classical methods for shallow bilinear-style networks. With efficient quantum access to the weights, we obtain a quartic speedup over classical methods. With efficient quantum access to both the inputs and the network weights, we prove that a network with an $N$-dimensional vectorized input, $k$ residual block layers, and a final residual-linear-pooling layer can be implemented with an error of $\epsilon$ with $O(\text{polylog}(N/\epsilon)^k)$ inference cost.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The manuscript presents the first fully-coherent quantum implementation of multilayer neural networks modeled on ResNet architectures, consisting of residual blocks with multi-filter 2D convolutions, sigmoid activations, skip connections, and layer normalizations. It analyzes inference complexity under three quantum data-access regimes: no assumptions (quadratic speedup for shallow bilinear networks), efficient quantum access to weights only (quartic speedup), and efficient quantum access to both inputs and weights (O(polylog(N/ε)^k) cost for an N-dimensional input, k residual-block layers, and a final residual-linear-pooling layer with error ε).

Significance. If the constructions and complexity bounds hold, the work would constitute a concrete advance in quantum machine learning by showing how non-linear activations and residual connections can be realized coherently, with explicit scaling that improves from quadratic to polylogarithmic under progressively stronger access models. The regime-based analysis is useful for clarifying the oracle requirements needed for quantum advantage in deep-network inference.

major comments (2)
  1. [§5] §5 (polylog regime): the O(polylog(N/ε)^k) bound is derived under the assumption of efficient coherent quantum oracles/QRAM for both inputs and all network weights, yet the manuscript provides no explicit query-complexity analysis or circuit construction showing how these oracles are realized while preserving superposition and coherence through the k layers, convolutions, sigmoid activations, and layer norms. Any super-logarithmic overhead in oracle implementation would invalidate the claimed scaling.
  2. [§4 and §5] §4 (quartic regime) and §5: the transition from the no-assumption quadratic regime to the weight-access quartic and then to the joint-access polylog regimes is load-bearing for the central claim, but the paper does not quantify the concrete gate or query overhead of implementing the residual blocks and non-linear functions under each access model, nor does it include an error-propagation analysis across layers.
minor comments (2)
  1. [Abstract] The abstract states the three regimes and final bound but omits any reference to the specific sections or equations containing the derivations; adding such pointers would improve readability.
  2. [§2] Notation for the vectorized input dimension N and the number of layers k is introduced without an explicit table or diagram summarizing the network architecture parameters used in the complexity statements.

Simulated Author's Rebuttal

2 responses · 0 unresolved

Thank you for your thorough review and for recognizing the potential significance of our work on coherent quantum implementations of multilayer neural networks. We address each of the major comments in detail below and have made revisions to the manuscript to incorporate clarifications and additional analyses as suggested.

read point-by-point responses

  1. Referee: [§5] §5 (polylog regime): the O(polylog(N/ε)^k) bound is derived under the assumption of efficient coherent quantum oracles/QRAM for both inputs and all network weights, yet the manuscript provides no explicit query-complexity analysis or circuit construction showing how these oracles are realized while preserving superposition and coherence through the k layers, convolutions, sigmoid activations, and layer norms. Any super-logarithmic overhead in oracle implementation would invalidate the claimed scaling.

    Authors: We thank the referee for this observation. Our polylogarithmic bound is stated under the standard assumption of efficient coherent quantum oracles (QRAM-style) for inputs and weights, which by definition support O(polylog N) queries while preserving superposition. The network constructions (convolutions via quantum linear algebra techniques, sigmoid via coherent polynomial approximations, layer norms via quantum mean/variance estimation, and residual additions) are composed to interface directly with these oracles. To make the query complexity and coherence preservation explicit, we have added a dedicated subsection in §5 together with an appendix that tabulates the per-component oracle queries and shows that no super-logarithmic overhead is introduced beyond the factors already absorbed in the O(polylog(N/ε)^k) expression. revision: yes

  2. Referee: [§4 and §5] §4 (quartic regime) and §5: the transition from the no-assumption quadratic regime to the weight-access quartic and then to the joint-access polylog regimes is load-bearing for the central claim, but the paper does not quantify the concrete gate or query overhead of implementing the residual blocks and non-linear functions under each access model, nor does it include an error-propagation analysis across layers.

    Authors: We agree that explicit overhead quantification and error propagation strengthen the regime transitions. In the revised manuscript we have expanded §4 and §5 with (i) concrete gate-count estimates for each residual-block primitive (2-D convolution, sigmoid, layer norm, skip connection) under the three access models, and (ii) a new error-propagation subsection showing that choosing per-layer precision O(ε/k) keeps the total accumulated error ≤ ε while preserving the stated complexity scalings. These additions make the quartic-to-polylog transition fully rigorous. revision: yes

Circularity Check

0 steps flagged

No circularity; complexity bounds derived from standard quantum primitives under external access assumptions

full rationale

The paper derives quadratic, quartic, and polylog(N/ε)^k inference costs for ResNet-style networks by composing known quantum algorithms (block-encodings, quantum linear algebra for convolutions, and coherent implementations of sigmoid/layer-norm) under three explicitly stated regimes of quantum data access. The O(polylog(N/ε)^k) claim for k residual blocks follows directly from the assumed efficient oracles/QRAM without any reduction of a derived quantity to a fitted parameter, self-defined term, or unverified self-citation chain. All steps remain self-contained against external quantum query complexity benchmarks and do not invoke internal fits or renamings that would create circularity.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The central claims rest on standard quantum computing assumptions about coherent operations and data access oracles rather than on new free parameters or invented entities.

axioms (2)
  • domain assumption Efficient quantum access to inputs and weights is possible via oracles or QRAM without destroying coherence
    Invoked to obtain the quartic and polylog speedups; stated in the abstract when describing the three regimes.
  • domain assumption Non-linear activations such as sigmoid can be implemented coherently on quantum states with controlled error
    Required for the multilayer construction to remain fully coherent.

pith-pipeline@v0.9.0 · 5745 in / 1550 out tokens · 36145 ms · 2026-05-21T20:54:41.666399+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

134 extracted references · 134 canonical work pages · 6 internal anchors

  1. [1]

    Deep learning,

    Yann LeCun, Yoshua Bengio, and Geoffrey Hinton. Deep learning.Nature, 521(7553): 436–444, 2015. URLhttps://doi.org/10.1038/nature14539

    work page doi:10.1038/nature14539 2015

  2. [2]

    MIT Press, 2016

    Ian Goodfellow, Yoshua Bengio, and Aaron Courville.Deep Learning. MIT Press, 2016. URL http://www.deeplearningbook.org

    work page 2016

  3. [3]

    2016.280

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition. InProceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pages 770–778, 2016. URLhttps://doi.org/10.1109/CVPR. 2016.90

    work page doi:10.1109/cvpr 2016

  4. [4]

    Denoising diffusion probabilistic models

    Jonathan Ho, Ajay Jain, and Pieter Abbeel. Denoising diffusion probabilistic models. In H. Larochelle, M. Ranzato, R. Hadsell, M.F. Balcan, and H. Lin, editors,Advances in Neural Information Processing Systems, volume 33, pages 6840–6851. Curran Associates, Inc.,

    work page

  5. [5]

    URL https://proceedings.neurips.cc/paper_files/paper/2020/ hash/4c5bcfec8584af0d967f1ab10179ca4b-Abstract.html

    work page 2020

  6. [6]

    An image is worth 16x16 words: Transformers for image recognition at scale

    Alexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn, Xiaohua Zhai, Thomas Unterthiner, Mostafa Dehghani, Matthias Minderer, Georg Heigold, Sylvain Gelly, Jakob Uszkoreit, and Neil Houlsby. An image is worth 16x16 words: Transformers for image recognition at scale. InInternational Conference on Learning Representations, 2021. URL https:...

    work page 2021

  7. [7]

    Attention is all you need

    Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, Ł ukasz Kaiser, and Illia Polosukhin. Attention is all you need. In I. Guyon, U. V on Luxburg, S. Bengio, H. Wallach, R. Fergus, S. Vishwanathan, and R. Garnett, editors, Advances in Neural Information Processing Systems, volume 30. Curran Associates, Inc.,

    work page

  8. [8]

    URL https://proceedings.neurips.cc/paper_files/paper/2017/ hash/3f5ee243547dee91fbd053c1c4a845aa-Abstract.html

    work page 2017

  9. [9]

    Language models are few-shot learners

    Tom Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah, Jared D Kaplan, Prafulla Dhari- wal, Arvind Neelakantan, Pranav Shyam, Girish Sastry, Amanda Askell, Sandhini Agar- wal, Ariel Herbert-V oss, Gretchen Krueger, Tom Henighan, Rewon Child, Aditya Ramesh, Daniel Ziegler, Jeffrey Wu, Clemens Winter, Chris Hesse, Mark Chen, Eric Sigler, Mateusz Litwin, Sco...

    work page 1901

  10. [10]

    URL https://proceedings.neurips.cc/paper_files/paper/2020/ hash/1457c0d6bfcb4967418bfb8ac142f64a-Abstract.html

    work page 2020

  11. [11]

    David Silver, Aja Huang, Chris J. Maddison, Arthur Guez, Laurent Sifre, George van den Driessche, Julian Schrittwieser, Ioannis Antonoglou, Veda Panneershelvam, Marc Lanctot, Sander Dieleman, Dominik Grewe, John Nham, Nal Kalchbrenner, Ilya Sutskever, Timothy Lillicrap, Madeleine Leach, Koray Kavukcuoglu, Thore Graepel, and Demis Hassabis. Master- ing the...

    work page doi:10.1038/nature16961 2016

  12. [12]

    John Jumper, Richard Evans, Alexander Pritzel, Tim Green, Michael Figurnov, Olaf Ron- neberger, Kathryn Tunyasuvunakool, Russ Bates, Augustin Žídek, Anna Potapenko, Alex Bridgland, Clemens Meyer, Simon A. A. Kohl, Andrew J. Ballard, Andrew Cowie, Bernardino Romera-Paredes, Stanislav Nikolov, Rishub Jain, Jonas Adler, Trevor Back, Stig Petersen, David Reim...

    work page doi:10.1038/s41586-021-03819-2 2021

  13. [13]

    Alhussein Fawzi, Matej Balog, Aja Huang, Thomas Hubert, Bernardino Romera-Paredes, Mohammadamin Barekatain, Alexander Novikov, Francisco J. R. Ruiz, Julian Schrit- twieser, Grzegorz Swirszcz, David Silver, Demis Hassabis, and Pushmeet Kohli. Discov- ering faster matrix multiplication algorithms with reinforcement learning.Nature, 610 (7930):47–53, October...

    work page 2022

  14. [14]

    Gordon E. Moore. Cramming more components onto integrated circuits.Electronics, 38(8):114–117, April 1965. URL https://ieeexplore.ieee.org/document/ 4785860

    work page 1965

  15. [15]

    Richard P. Feynman. Simulating physics with computers.International Journal of Theoretical Physics, 21(6–7):467–488, June 1982. ISSN 1572-9575. URL https://doi.org/10. 1007/BF02650179

    work page 1982

  16. [16]

    Richard P. Feynman. Quantum mechanical computers.Foundations of Physics, 16(6):507–531, June 1986. ISSN 1572-9516. URLhttps://doi.org/10.1007/BF01886518

    work page doi:10.1007/bf01886518 1986

  17. [17]

    Nielsen and Isaac L

    Michael A. Nielsen and Isaac L. Chuang.Quantum Computation and Quantum Information. Cambridge University Press, 2010. ISBN 9781107002173. URL https://doi.org/10. 1017/CBO9780511976667

    work page 2010

  18. [18]

    [Online]

    Jacob Biamonte, Peter Wittek, Nicola Pancotti, Patrick Rebentrost, Nathan Wiebe, and Seth Lloyd. Quantum machine learning.Nature, 549(7671):195–202, 2016. URL https: //doi.org/10.1038/nature23474

    work page doi:10.1038/nature23474 2016

  19. [19]

    Schuld and F

    M. Schuld and F. Petruccione.Machine Learning with Quantum Computers. Springer Cham, 2021. ISBN 9783030830984. URL https://doi.org/10.1007/ 978-3-030-83098-4

    work page 2021

  20. [20]

    Quantum machine learning: A hands-on tutorial for machine learning practitioners and researchers, 2025

    Yuxuan Du, Xinbiao Wang, Naixu Guo, Zhan Yu, Yang Qian, Kaining Zhang, Min-Hsiu Hsieh, Patrick Rebentrost, and Dacheng Tao. Quantum machine learning: A hands-on tutorial for machine learning practitioners and researchers, 2025. URL https://arxiv.org/abs/ 2502.01146

    work page arXiv 2025

  21. [21]

    Quantum Computing in the NISQ era and beyond

    John Preskill. Quantum computing in the NISQ era and beyond.Quantum, 2:79, August 2018. URLhttps://doi.org/10.22331/q-2018-08-06-79. 11

    work page internal anchor Pith review doi:10.22331/q-2018-08-06-79 2018

  22. [22]

    Nature Communications , author =

    Alberto Peruzzo, Jarrod McClean, Peter Shadbolt, Man-Hong Yung, Xiao-Qi Zhou, Peter J. Love, Alán Aspuru-Guzik, and Jeremy L. O’Brien. A variational eigenvalue solver on a photonic quantum processor.Nature Communications, 5:4213, 2014. URLhttps://doi. org/10.1038/ncomms5213

    work page doi:10.1038/ncomms5213 2014

  23. [23]

    Variational Quantum Algorithms,

    Marco Cerezo, Andrew Arrasmith, Ryan Babbush, Simon C. Benjamin, Suguru Endo, Keisuke Fujii, Jarrod R. McClean, Kosuke Mitarai, Xiao Yuan, Lukasz Cincio, and Patrick J. Coles. Variational quantum algorithms.Nat. Rev. Phys, 3(9):625–644, August 2021. URL http: //doi.org/10.1038/s42254-021-00348-9

    work page doi:10.1038/s42254-021-00348-9 2021

  24. [24]

    Quantum Sci

    Marcello Benedetti, Erika Lloyd, Stefan Sack, and Mattia Fiorentini. Parameterized quantum circuits as machine learning models.Quantum Sci. Technol., 4(4):043001, November 2019. URLhttp://doi.org/10.1088/2058-9565/ab4eb5

    work page doi:10.1088/2058-9565/ab4eb5 2019

  25. [25]

    Training variational quantum algorithms is NP-hard.Phys

    Lennart Bittel and Martin Kliesch. Training variational quantum algorithms is NP-hard.Phys. Rev. Lett., 127:120502, Sep 2021. URL https://link.aps.org/doi/10.1103/ PhysRevLett.127.120502

    work page 2021

  26. [26]

    2022 , month =

    Eric R. Anschuetz and Bobak T. Kiani. Quantum variational algorithms are swamped with traps.Nature Communications, 13(1), December 2022. ISSN 2041-1723. URL https: //doi.org/10.1038/s41467-022-35364-5

    work page doi:10.1038/s41467-022-35364-5 2022

  27. [27]

    2018 , month =

    Jarrod R. McClean, Sergio Boixo, Vadim N. Smelyanskiy, Ryan Babbush, and Hartmut Neven. Barren plateaus in quantum neural network training landscapes.Nature Communications, 9 (1), November 2018. URLhttps://doi.org/10.1038/s41467-018-07090-4

    work page doi:10.1038/s41467-018-07090-4 2018

  28. [28]

    Coles, Lukasz Cincio, Jarrod R

    Martín Larocca, Supanut Thanasilp, Samson Wang, Kunal Sharma, Jacob Biamonte, Patrick J. Coles, Lukasz Cincio, Jarrod R. McClean, Zoë Holmes, and M. Cerezo. Barren plateaus in variational quantum computing, March 2025. ISSN 2522-5820. URL https://doi.org/ 10.1038/s42254-025-00813-9

    work page doi:10.1038/s42254-025-00813-9 2025

  29. [29]

    Cerezo, Martin Larocca, Diego García-Martín, N

    M. Cerezo, Martin Larocca, Diego García-Martín, N. L. Diaz, Paolo Braccia, Enrico Fontana, Manuel S. Rudolph, Pablo Bermejo, Aroosa Ijaz, Supanut Thanasilp, Eric R. Anschuetz, and Zoë Holmes. Does provable absence of barren plateaus imply classical simulability?Nature Communications, 16(1), August 2025. ISSN 2041-1723. URL https://doi.org/10. 1038/s41467-...

    work page 2025

  30. [30]

    Rudolph, Zoë Holmes, Lukasz Cincio, and M

    Pablo Bermejo, Paolo Braccia, Manuel S. Rudolph, Zoë Holmes, Lukasz Cincio, and M. Cerezo. Quantum convolutional neural networks are (effectively) classically simulable,

    work page

  31. [31]

    URLhttps://arxiv.org/abs/2408.12739

    work page internal anchor Pith review Pith/arXiv arXiv

  32. [32]

    doi:10.1038/s41586-019-0980-2

    V ojtˇech Havlíˇcek, Antonio D. Córcoles, Kristan Temme, Aram W. Harrow, Abhinav Kandala, Jerry M. Chow, and Jay M. Gambetta. Supervised learning with quantum-enhanced feature spaces.Nature, 567(7747):209–212, March 2019. ISSN 1476-4687. URL https://doi. org/10.1038/s41586-019-0980-2

    work page doi:10.1038/s41586-019-0980-2 2019

  33. [33]

    Quantum machine learning in feature Hilbert spaces

    Maria Schuld and Nathan Killoran. Quantum machine learning in feature Hilbert spaces. Phys. Rev. Lett., 122:040504, February 2019. URL https://link.aps.org/doi/10. 1103/PhysRevLett.122.040504

    work page 2019

  34. [34]

    A generative modeling approach for benchmarking and training shallow quantum circuits.npj Quantum Information, 5(1), May 2019

    Marcello Benedetti, Delfina Garcia-Pintos, Oscar Perdomo, Vicente Leyton-Ortega, Yunseong Nam, and Alejandro Perdomo-Ortiz. A generative modeling approach for benchmarking and training shallow quantum circuits.npj Quantum Information, 5(1), May 2019. ISSN 2056-6387. URLhttps://doi.org/10.1038/s41534-019-0157-8

    work page doi:10.1038/s41534-019-0157-8 2019

  35. [35]

    Huang and P

    Po-Wei Huang and Patrick Rebentrost. Post-variational quantum neural networks, 2024. URL https://arxiv.org/abs/2307.10560

    work page arXiv 2024

  36. [36]

    Cerezo, and Zoë Holmes

    Supanut Thanasilp, Samson Wang, M. Cerezo, and Zoë Holmes. Exponential concentration in quantum kernel methods.Nature Communications, 15(1), June 2024. ISSN 2041-1723. URL https://doi.org/10.1038/s41467-024-49287-w. 12

    work page doi:10.1038/s41467-024-49287-w 2024

  37. [37]

    Rudolph, Sacha Lerch, Supanut Thanasilp, Oriel Kiss, Oxana Shaya, Sofia Val- lecorsa, Michele Grossi, and Zoë Holmes

    Manuel S. Rudolph, Sacha Lerch, Supanut Thanasilp, Oriel Kiss, Oxana Shaya, Sofia Val- lecorsa, Michele Grossi, and Zoë Holmes. Trainability barriers and opportunities in quantum generative modeling.npj Quantum Information, 10(1), November 2024. ISSN 2056-6387. URLhttps://doi.org/10.1038/s41534-024-00902-0

    work page doi:10.1038/s41534-024-00902-0 2024

  38. [38]

    Quantum A lgorithm for Linear Systems of Equations

    Aram W. Harrow, Avinatan Hassidim, and Seth Lloyd. Quantum algorithm for linear systems of equations.Phys. Rev. Lett., 103:150502, October 2009. URL https://link.aps. org/doi/10.1103/PhysRevLett.103.150502

    work page doi:10.1103/physrevlett.103.150502 2009

  39. [39]

    Quantum algorithms: an overview.npj Quantum Information, 2(1), January

    Ashley Montanaro. Quantum algorithms: an overview.npj Quantum Information, 2(1), January

    work page

  40. [40]

    Quantum algorithms: an overview,

    ISSN 2056-6387. URLhttps://doi.org/10.1038/npjqi.2015.23

    work page doi:10.1038/npjqi.2015.23 2056

  41. [41]

    Proceedings of the 51st Annual ACM SIGACT Symposium on Theory of Computing , pages =

    András Gilyén, Yuan Su, Guang Hao Low, and Nathan Wiebe. Quantum singular value transformation and beyond: exponential improvements for quantum matrix arithmetics. In Proceedings of the 51st Annual ACM SIGACT Symposium on Theory of Computing, STOC 2019, pages 193–204, New York, NY , USA, 2019. Association for Computing Machinery. ISBN 9781450367059. URLht...

    work page doi:10.1145/3313276.3316366 2019

  42. [42]

    Dalzell, Sam McArdle, Mario Berta, Przemyslaw Bienias, Chi-Fang Chen, András Gilyén, Connor T

    Alexander M. Dalzell, Sam McArdle, Mario Berta, Przemyslaw Bienias, Chi-Fang Chen, András Gilyén, Connor T. Hann, Michael J. Kastoryano, Emil T. Khabiboulline, Aleksander Kubica, Grant Salton, Samson Wang, and Fernando G. S. L. Brandão. Quantum algorithms: A survey of applications and end-to-end complexities, April 2025. URL https://doi.org/ 10.1017/9781009639651

    work page doi:10.1017/9781009639651 2025

  43. [43]

    & Lloyd, S

    Patrick Rebentrost, Masoud Mohseni, and Seth Lloyd. Quantum support vector machine for big data classification.Phys. Rev. Lett., 113:130503, September 2014. URL https: //link.aps.org/doi/10.1103/PhysRevLett.113.130503

    work page doi:10.1103/physrevlett.113.130503 2014

  44. [44]

    Quantum algorithm for data fitting.Phys

    Nathan Wiebe, Daniel Braun, and Seth Lloyd. Quantum algorithm for data fitting.Phys. Rev. Lett., 109:050505, Aug 2012. URL https://link.aps.org/doi/10.1103/ PhysRevLett.109.050505

    work page 2012

  45. [45]

    Quantum algorithms for feedforward neural networks.ACM Trans

    Jonathan Allcock, Chang-Yu Hsieh, Iordanis Kerenidis, and Shengyu Zhang. Quantum algorithms for feedforward neural networks.ACM Trans. Quantum Comput., 1(1), October

    work page

  46. [46]

    URLhttps://doi.org/10.1145/3411466

    work page doi:10.1145/3411466

  47. [47]

    Quantum algorithms for deep convolutional neural networks

    Iordanis Kerenidis, Jonas Landman, and Anupam Prakash. Quantum algorithms for deep convolutional neural networks. InInternational Conference on Learning Representations,

    work page

  48. [48]

    URLhttps://openreview.net/forum?id=Hygab1rKDS

    work page

  49. [49]

    Quantum linear algebra is all you need for Transformer architectures,

    Naixu Guo, Zhan Yu, Matthew Choi, Aman Agrawal, Kouhei Nakaji, Alán Aspuru-Guzik, and Patrick Rebentrost. Quantum linear algebra is all you need for Transformer architectures,

    work page

  50. [50]

    URLhttps://arxiv.org/abs/2402.16714

    work page arXiv

  51. [51]

    Quantum principal component analysis

    Seth Lloyd, Masoud Mohseni, and Patrick Rebentrost. Quantum principal component analysis. Nature Physics, 10(9):631–633, July 2014. ISSN 1745-2481. URL https://doi.org/ 10.1038/nphys3029

    work page doi:10.1038/nphys3029 2014

  52. [52]

    Nathan Wiebe, Ashish Kapoor, and Krysta M. Svore. Quantum deep learning.Quantum Information and Computation, 16(7 & 8):541–587, May 2016. ISSN 1533-7146. URL https://doi.org/10.26421/QIC16.7-8-1

    work page doi:10.26421/qic16.7-8-1 2016

  53. [53]

    Bromley, Christian Weedbrook, and Seth Lloyd

    Patrick Rebentrost, Thomas R. Bromley, Christian Weedbrook, and Seth Lloyd. Quantum Hopfield neural network.Phys. Rev. A, 98:042308, October 2018. URL https://link. aps.org/doi/10.1103/PhysRevA.98.042308

    work page doi:10.1103/physreva.98.042308 2018

  54. [54]

    Quantum perceptron models

    Ashish Kapoor, Nathan Wiebe, and Krysta Svore. Quantum perceptron models. In D. Lee, M. Sugiyama, U. Luxburg, I. Guyon, and R. Garnett, editors,Ad- vances in Neural Information Processing Systems, volume 29. Curran Associates, Inc., 2016. URL https://proceedings.neurips.cc/paper/2016/hash/ d47268e9db2e9aa3827bba3afb7ff94a-Abstract.html. 13

    work page 2016

  55. [55]

    Quantum vision transformers.Quantum, 8:1265, February 2024

    El Amine Cherrat, Iordanis Kerenidis, Natansh Mathur, Jonas Landman, Martin Strahm, and Yun Yvonna Li. Quantum vision transformers.Quantum, 8:1265, February 2024. ISSN 2521-327X. URLhttps://doi.org/10.22331/q-2024-02-22-1265

    work page doi:10.22331/q-2024-02-22-1265 2024

  56. [56]

    A rig orous and robust quantum speed-up in supervised machine learning

    Yunchao Liu, Srinivasan Arunachalam, and Kristan Temme. A rigorous and robust quantum speed-up in supervised machine learning.Nature Physics, 17:1013–1017, 2021. URL https: //doi.org/10.1038/s41567-021-01287-z

    work page doi:10.1038/s41567-021-01287-z 2021

  57. [57]

    Quantum Alphatron: quantum advantage for learning with kernels and noise.Quantum, 7:1174, November 2023

    Siyi Yang, Naixu Guo, Miklos Santha, and Patrick Rebentrost. Quantum Alphatron: quantum advantage for learning with kernels and noise.Quantum, 7:1174, November 2023. ISSN 2521-327X. URLhttps://doi.org/10.22331/q-2023-11-08-1174

    work page doi:10.22331/q-2023-11-08-1174 2023

  58. [58]

    QKAN: Quantum Kolmogorov-Arnold networks, 2024

    Petr Ivashkov, Po-Wei Huang, Kelvin Koor, Lirandë Pira, and Patrick Rebentrost. QKAN: Quantum Kolmogorov-Arnold networks, 2024. URL https://arxiv.org/abs/2410. 04435

    work page 2024

  59. [59]

    Towards efficient quantum algorithms for diffusion probability models, 2025

    Yunfei Wang, Ruoxi Jiang, Yingda Fan, Xiaowei Jia, Jens Eisert, Junyu Liu, and Jin-Peng Liu. Towards efficient quantum algorithms for diffusion probability models, 2025. URL https://arxiv.org/abs/2502.14252

    work page arXiv 2025

  60. [60]

    Quantum gradient descent for linear systems and least squares.Phys

    Iordanis Kerenidis and Anupam Prakash. Quantum gradient descent for linear systems and least squares.Phys. Rev. A, 101(2), February 2020. ISSN 2469-9934. URL https: //link.aps.org/doi/10.1103/PhysRevA.101.022316

    work page doi:10.1103/physreva.101.022316 2020

  61. [61]

    Huggins, Ramis Movas- sagh, Dar Gilboa, and Jarrod McClean

    Amira Abbas, Robbie King, Hsin-Yuan Huang, William J. Huggins, Ramis Movas- sagh, Dar Gilboa, and Jarrod McClean. On quantum backpropagation, informa- tion reuse, and cheating measurement collapse. In A. Oh, T. Naumann, A. Glober- son, K. Saenko, M. Hardt, and S. Levine, editors,Advances in Neural Informa- tion Processing Systems, volume 36, pages 44792–4...

    work page

  62. [62]

    URL https://proceedings.neurips.cc/paper_files/paper/2023/ hash/8c3caae2f725c8e2a55ecd600563d172-Abstract.html

    work page 2023

  63. [63]

    Towards provably efficient quantum algorithms for large-scale machine- learning models.Nature Communications, 15:434, 2024

    Ji Liu, Mengzhen Liu, Jiapeng Liu, Chengran Peng, Liyuan Zhang, Xiao Yuan, Peng Xue, and Xiongfeng Ma. Towards provably efficient quantum algorithms for large-scale machine- learning models.Nature Communications, 15:434, 2024. URL https://doi.org/10. 1038/s41467-023-43957-x

    work page 2024

  64. [64]

    Aggregated residual transformations for deep neural networks

    Saining Xie, Ross Girshick, Piotr Dollár, Zhuowen Tu, and Kaiming He. Aggregated residual transformations for deep neural networks. In2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pages 5987–5995, 2017. URL https://doi.org/10. 1109/CVPR.2017.634

    work page 2017

  65. [65]

    Attention is not all you need: pure attention loses rank doubly exponentially with depth

    Yihe Dong, Jean-Baptiste Cordonnier, and Andreas Loukas. Attention is not all you need: pure attention loses rank doubly exponentially with depth. In Marina Meila and Tong Zhang, editors,Proceedings of the 38th International Conference on Machine Learning, volume 139 ofProceedings of Machine Learning Research, pages 2793–2803. PMLR, 18–24 Jul 2021. URLhtt...

    work page 2021

  66. [66]

    Iris Cong, Soonwon Choi, and Mikhail D. Lukin. Quantum convolutional neural networks. Nature Physics, 15(12):1273–1278, August 2019. URL http://doi.org/10.1038/ s41567-019-0648-8

    work page 2019

  67. [67]

    Rattew and Patrick Rebentrost

    Arthur G. Rattew and Patrick Rebentrost. Non-linear transformations of quantum amplitudes: Exponential improvement, generalization, and applications, 2023. URL https://arxiv. org/abs/2309.09839

    work page arXiv 2023

  68. [68]

    Model compression and acceleration for deep neural networks: The principles, progress, and challenges.IEEE Signal Processing Magazine, 35(1):126–136, 2018

    Yu Cheng, Duo Wang, Pan Zhou, and Tao Zhang. Model compression and acceleration for deep neural networks: The principles, progress, and challenges.IEEE Signal Processing Magazine, 35(1):126–136, 2018. URLhttps://doi.org/10.1109/MSP.2017.2765695

    work page doi:10.1109/msp.2017.2765695 2018

  69. [69]

    Wide residual networks

    Sergey Zagoruyko and Nikos Komodakis. Wide residual networks. InProcedings of the British Machine Vision Conference 2016, BMVC 2016, pages 87.1–87.12. British Machine Vision Association, 2016. URLhttps://doi.org/10.5244/C.30.87. 14

    work page doi:10.5244/c.30.87 2016

  70. [70]

    Samuel Jaques and Arthur G. Rattew. QRAM: A survey and critique, 2023. URL https: //arxiv.org/abs/2305.10310

    work page arXiv 2023

  71. [71]

    Dalzell, András Gilyén, Connor T

    Alexander M. Dalzell, András Gilyén, Connor T. Hann, Sam McArdle, Grant Salton, Quynh T. Nguyen, Aleksander Kubica, and Fernando G. S. L. Brandão. A distillation-teleportation pro- tocol for fault-tolerant QRAM, 2025. URLhttps://arxiv.org/abs/2505.20265

    work page arXiv 2025

  72. [72]

    Bilinear CNN models for fine- grained visual recognition

    Tsung-Yu Lin, Aruni RoyChowdhury, and Subhransu Maji. Bilinear CNN models for fine- grained visual recognition. In2015 IEEE International Conference on Computer Vision (ICCV), pages 1449–1457. IEEE, December 2015. URL https://doi.org/10.1109/ ICCV.2015.170

    work page 2015

  73. [73]

    Quantum-state preparation with universal gate decomposi- tions.Phys

    Martin Plesch and ˇCaslav Brukner. Quantum-state preparation with universal gate decomposi- tions.Phys. Rev. A, 83:032302, March 2011. URL https://link.aps.org/doi/10. 1103/PhysRevA.83.032302

    work page 2011

  74. [74]

    Quan- tum algorithm and circuit design solving the Poisson equation.New Journal of Physics, 15(1):013021, January 2013

    Yudong Cao, Anargyros Papageorgiou, Iasonas Petras, Joseph Traub, and Sabre Kais. Quan- tum algorithm and circuit design solving the Poisson equation.New Journal of Physics, 15(1):013021, January 2013. ISSN 1367-2630. URL https://doi.org/10.1088/ 1367-2630/15/1/013021

    work page 2013

  75. [75]

    High-precision quantum algorithms for partial differential equations.Quantum, 5:574, 2021

    Andrew M Childs, Jin-Peng Liu, and Aaron Ostrander. High-precision quantum algorithms for partial differential equations.Quantum, 5:574, 2021. URL https://doi.org/10. 22331/q-2021-11-10-574

    work page 2021

  76. [76]

    Berry and Pedro C

    Dominic W. Berry and Pedro C. S. Costa. Quantum algorithm for time-dependent differential equations using Dyson series.Quantum, 8:1369, June 2024. ISSN 2521-327X. URL https://doi.org/10.22331/q-2024-06-13-1369

    work page doi:10.22331/q-2024-06-13-1369 2024

  77. [77]

    The cost of solving linear differential equations on a quantum computer: fast-forwarding to explicit resource counts.Quantum, 8:1553, 2024

    David Jennings, Matteo Lostaglio, Robert B Lowrie, Sam Pallister, and Andrew T Sornborger. The cost of solving linear differential equations on a quantum computer: fast-forwarding to explicit resource counts.Quantum, 8:1553, 2024. URL https://doi.org/10.22331/ q-2024-12-10-1553

    work page 2024

  78. [78]

    Fast-forwarding quantum algorithms for linear dissipative differential equations, 2024

    Dong An, Akwum Onwunta, and Gengzhi Yang. Fast-forwarding quantum algorithms for linear dissipative differential equations, 2024. URL https://arxiv.org/abs/2410. 13189

    work page 2024

  79. [79]

    Designing a nearly optimal quantum algorithm for linear differential equations via Lindbladians.Phys

    Zhong-Xia Shang, Naixu Guo, Dong An, and Qi Zhao. Designing a nearly optimal quantum algorithm for linear differential equations via Lindbladians.Phys. Rev. Lett., 135:120604, Sep

    work page

  80. [80]

    URLhttps://link.aps.org/doi/10.1103/cvl9-97qg

    work page doi:10.1103/cvl9-97qg

Showing first 80 references.