Pith. sign in

REVIEW 3 major objections 5 minor 2 cited by

This paper claims that a two-cavity atom neuron can implement an all-optical, tunable nonlinear activation function for optical neural networks, delivering over 95 percent accuracy on MNIST and satellite-image classification without hidden-

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

T0 review · deepseek-v4-flash

2026-08-03 23:19 UTC pith:HMZDDRRH

load-bearing objection The central nonlinearity is built on a questionable mapping from atomic population to coherent photon amplitude, so the simulated accuracies describe a classical activation rather than the proposed quantum hardware. the 3 major comments →

arxiv 2511.06167 v2 pith:HMZDDRRH submitted 2025-11-09 quant-ph physics.atom-phphysics.optics

Quantum optical neural networks using atom-cavity interactions to provide all-optical nonlinearity

Chuanzhou Zhu , Tianyu Wang , Peter L. McMahon , Daniel Soh This is my paper
classification quant-ph physics.atom-phphysics.optics
keywords quantum optical neural networkatom-cavity neuronall-optical nonlinearitycavity QEDnonlinear activation functionMNIST classificationSAT-6 remote sensingphoton loss robustness
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

The paper proposes an optical neural network whose hidden layers are made of atom-cavity 'neurons' that absorb and re-emit single photons, replacing the electronic detectors and emitters normally needed for nonlinear activation. The central claim is that a two-level atom coupled to two switchable cavities produces an activation function—a = (g|z|/Ω)|sin(πtΩ)|, with Ω = sqrt((gz)² + δ²)—whose shape is tuned by the photon absorption time t. If this holds, the network can carry out inference entirely in the optical domain, with weights set by spatial-light-modulator pixels and trained by ordinary backpropagation. Simulated on MNIST and the SAT-6 satellite imagery benchmark, the network reaches over 95 percent accuracy and stays robust to random detuning and photon loss, suggesting a low-power path to onboard satellite sensing.

Core claim

The discovery is that a single two-level atom in a two-cavity neuron can serve as a differentiable, tunable nonlinearity for optical neural networks. During absorption, the atom undergoes Rabi oscillation at rate Ω = sqrt((gz)² + δ²), so after an interaction time t the excitation, and hence the emitted photon intensity, is a sinusoidal function of the incoming amplitude z. With a weak auxiliary laser to lock the spontaneous-emission phase, the activation becomes a deterministic, real-valued, differentiable function of z, enabling gradient-based training. The authors show that varying the absorption time t balances nonlinearity against monotonicity, that random per-neuron detuning is tolerate

What carries the argument

The central object is the atom-cavity neuron: a two-level atom placed between a low-Q cavity (for absorption) and a switchable high-Q cavity (for storage and emission). Its activation function is the Rabi-oscillation expression a(z) = g|z|/Ω · |sin(π t Ω)| with Ω = sqrt((gz)² + δ²), where t is the photon absorption duration, g the atom-photon coupling, and δ the atom-cavity detuning. This function carries the argument: it is the all-optical nonlinearity that makes the network trainable and lets hidden layers function without photodetection or re-emission electronics.

Load-bearing premise

The two-step physical assumption that an absorbed photon's atomic excitation converts with unit efficiency into an emitted photon of intensity (1/2)(⟨σz⟩+1), and that a weak auxiliary laser fully locks the spontaneous-emission phase, is asserted without derivation; if either fails, the simulated classification numbers do not reflect what the hardware would produce.

What would settle it

Measure the activation of a single two-cavity neuron: drive it with coherent pulses of controlled amplitude z and absorption time t, and record the emitted photon intensity against the prediction |a|² = (g|z|/(2Ω))(1 − cos(2π t Ω)) plus the phase-locking assumption. A deviation in the oscillation envelope, or a residual random phase that cannot be compensated by the auxiliary laser, would break Eq. (7) and invalidate the simulated network accuracy.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

If this is right

  • Hidden layers of an optical neural network can, in principle, run without any electronic photon detection or emission, removing two major sources of latency and energy draw.
  • The absorption time t acts as a tunable, layer-specific control knob for network nonlinearity; optimal performance occurs at an intermediate value where the activation is neither too linear nor too oscillatory.
  • The network tolerates large random atom-cavity detuning and photon pass rates as low as 20 percent (still ~80 percent MNIST accuracy), suggesting that fabrication imprecision need not be fatal.
  • A convolutional QONN achieves comparable SAT-6 accuracy while cutting the number of SLM-controlled parameters by orders of magnitude (from over five million to as few as 150), pointing toward compact onboard systems.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • If the phase-locking step behaves as assumed, the same differentiable activation could slot into other photonic learning architectures that currently rely on optoelectronic nonlinearities, though the paper does not explore those.
  • The robustness to random detuning suggests that manufacturing variation across many cavity neurons could be exploited as a built-in diversity mechanism rather than corrected, which could relax fabrication tolerances in solid-state implementations.
  • The mean-field approximation neglects photon entanglement created in the optical multiplier; a fully quantum treatment could change the effective activation and either improve or degrade accuracy, a direct testable follow-up.
  • Because the activation output is nonnegative and bounded in [0,1], the network cannot represent negative values directly; adding per-neuron biases or differential (push-pull) neuron pairs would be a natural extension to test expressivity limits.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. The paper proposes a quantum optical neural network in which atom-cavity 'neurons' provide the nonlinear activation between optical matrix-vector multiplications. Each neuron absorbs an incident photon amplitude z_i^l, the two-level atom undergoes a Rabi oscillation (Eqs. (1)-(2)), and the atom is then claimed to re-emit a photon whose intensity equals the atomic excitation probability (Eq. (4)); after phase-locking by an auxiliary laser, this yields the scalar activation function a_i^l = g|z_i^l|/Omega_i^l |sin(pi t^l Omega_i^l)| (Eq. (7)). The authors simulate MNIST classification with two fully-connected hidden layers, study the effects of absorption time, random detuning, and stochastic photon loss, and introduce a convolutional variant applied to the SAT-6 satellite imagery task, reporting >95% accuracy in both cases.

Significance. If Eq. (7) were a faithful input-output relation of the proposed cavity-QED hardware, the proposal would be significant: it would replace hidden-layer photon detection/emission electronics in optical neural networks with tunable atom-cavity absorption, potentially enabling low-power onboard inference. The internal algebra from Eq. (2) to Eq. (7) is consistent, and the robustness studies (random detuning, stochastic loss) are clearly described and reproducible from the text. However, the central hardware claim rests on two unmodeled physical assumptions: that emitted field intensity equals atomic population (Eq. (4)) and that a weak auxiliary laser phase-locks the field without altering the dynamics. These assumptions are load-bearing because the subsequent MVM (Eq. (6)) requires interfering complex field amplitudes, and backpropagation (Appendix B) differentiates Eq. (7). The reported accuracies are therefore simulations of a scalar classical activation, not of the proposed quantum hardware, unless the missing quantum-optical derivation is supplied.

major comments (3)
  1. [Sec. II, Eq. (4)] Eq. (4) equates the emitted field intensity with the atomic population, |a|^2 = (1+<sigma_z>)/2. For a single two-level atom, the coherent field amplitude emitted into a cavity mode is proportional to the atomic coherence <sigma^->, not to the excited-state population. At the moment of complete excitation, <sigma_z>=+1 but the coherence vanishes, so Eq. (4) predicts maximal output while the deterministic coherent amplitude is zero. Since Eq. (6) combines complex field amplitudes by interference, the c-number activation a_i^l in Eq. (7) is not established. The authors need to derive the input-output relation from a cavity-QED master equation or input-output formalism, including the cavity field and atomic coherences.
  2. [Sec. II, Eq. (5) and phase-locking paragraph] The 'weak auxiliary external laser' is invoked to phase-lock the spontaneous-emission phase phi_i^l, allowing a, z, and W to be treated as real. A weak seed cannot define a phase reference for a single-photon-level field without contributing photons; a seed strong enough to establish a coherent output would modify the drive in Eq. (1) and therefore the activation. Moreover, if the neuron emits a one-photon Fock state, as implied by 'complete energy conversion,' the mean field <a> is zero, so the deterministic amplitude in Eq. (7) has no operational meaning without a local oscillator or heterodyne measurement. The manuscript should specify the seed strength, the measurement scheme, and show quantitatively that Eq. (7) survives.
  3. [App. B and Sec. III] The reported >95% MNIST and SAT-6 accuracies are obtained by training and testing the scalar activation of Eq. (7), not by simulating the proposed cavity-QED hardware. Because Eq. (7) is not yet derived from the physics, these results do not currently provide evidence that the hardware would achieve the claimed performance. The Discussion acknowledges mean-field neglect of entanglement, but it does not address the coherence/Fock-state issue. A full few-photon simulation or a coherent-state derivation with an explicit phase reference is needed before the central claim can be evaluated.
minor comments (5)
  1. [Sec. III.A] The text refers to 'the asymmetric colormap in Fig. 1a' when describing the t1-t2 accuracy map; this should be Fig. 2a.
  2. [Eq. (7)] The domain is stated as z_i^l in (-infinity, infinity), but the activation depends only on |z_i^l|, discarding the sign. The authors should comment on the expressivity implications or justify this choice.
  3. [Fig. 2c] The 'occurrence population distributions' of |z_i^1| and |z_i^2| are not defined precisely; please state how the histograms are computed (e.g., over the test set, over a single image).
  4. [Sec. II, phase-locking paragraph] Minor grammar: 'a weak auxiliary external laser need to be applied' should be 'needs to be applied.'
  5. [App. B] Typo: 'applys' should be 'applies.'

Circularity Check

0 steps flagged

No material circularity: the activation function is derived from the stated atom-cavity Hamiltonian, not imported from fitted outputs or from self-citation.

full rationale

The central derivation chain in Sec. II is self-contained. Eq. (2) is the Rabi solution of the Hamiltonian in Eq. (1); substituting Eq. (2) into Eq. (4) and taking the modulus (with the phase-locking assumption) gives Eq. (7), and for δ=0 it reduces to |sin(π t g |z|)|. No parameter appearing in Eq. (7) is fit to the MNIST/SAT-6 accuracies: g, δ, and t are stated physical parameters, and the network weights are trained by backpropagation on the training split. The benchmark simulations are self-consistent numerical evaluations of the proposed model, not a renaming of fitted outputs. The few self-citations ([75,76,79,80]) support auxiliary implementation or context points, and they do not carry the load of the activation derivation or the accuracy claims. Two caveats are correctness/validation concerns rather than circularity: Eq. (4) equates emitted photon intensity to the atomic excited-state population without deriving the coherence dynamics, and the absorption time t is selected using MNIST test accuracy in Fig. 2 rather than a held-out validation set. These affect whether the hardware would reproduce the simulated accuracy, but they do not reduce the derivation to its own inputs.

Axiom & Free-Parameter Ledger

4 free parameters · 6 axioms · 0 invented entities

No new fundamental entities (particles, forces, dimensions) are introduced; the atom-cavity neuron is an arrangement of known components. The main burdens are tunable simulation parameters (absorption time, detuning width, pass rate) and physical assumptions about emission conversion, phase-locking, and mean-field drive that are not experimentally or formally justified.

free parameters (4)
  • photon absorption times t1, t2 = t = 1 (with g = 1) chosen as optimal
    The activation function's nonlinearity/monotonicity depends on t; t is scanned over [0,4] and the best MNIST accuracy selects t≈1 (Fig. 2). This is a tuned hyperparameter, not derived from the hardware.
  • detuning distribution width δ0 = scanned up to 2.5
    Random detuning δ_i is drawn from U[-2δ0,2δ0]; δ0 is a simulation parameter used to test robustness, not measured from a real device.
  • photon pass rate P = P from 0 to 1 (Fig. 4)
    Stochastic photon loss is modeled by a Bernoulli pass rate P; it is a scanned parameter, not a measured loss.
  • coupling strength g = g = 1 in all simulations
    g sets the frequency scale and appears with t as the product t*g; setting g=1 is a normalization rather than an independently fitted physical constant. Included for completeness because Eq. (7) uses it explicitly.
axioms (6)
  • domain assumption The incident photon field can be treated as a classical mean-field drive z, with the atom undergoing Rabi oscillations (Eqs. 1-3).
    The paper explicitly uses a mean-field treatment and neglects photon entanglement; single-photon amplitude z is treated as a coherent drive.
  • ad hoc to paper After excitation, the high-Q cavity converts atomic excitation to a photon with intensity |a|^2=(1+⟨σz⟩)/2 (Eq. 4).
    This 'complete energy conversion' relation is stated without derivation and is essential to obtain the activation function.
  • ad hoc to paper A weak auxiliary external laser phase-locks the spontaneous emission phase, letting a, z, and W be treated as real (Sec. II after Eq. 6).
    Random spontaneous phase φ_i in Eq. (5) is eliminated by an asserted phase-locking mechanism that is not modeled; without it Eq. (7) is not valid.
  • domain assumption The high-Q cavity can be switched on/off and holds the excitation without photon emission during Step 2.
    Temporal multiplexing and synchronized layer emission require that the high-Q cavity stores excitation between absorption and emission; this storage fidelity is assumed.
  • domain assumption Photon loss can be modeled as a Bernoulli mask in the forward pass and by a mean-field scaling in the backward pass (Appendix C).
    The stochastic loss model is a semi-classical approximation; true few-photon loss would be a quantum jump process.
  • domain assumption Weights W are constrained to [-1,1] and are real; an SLM can implement the trained real-valued weights with negligible error.
    The optical MVM/SLM is assumed to implement the trained weights without calibration error or optical noise; no SLM noise is modeled.

pith-pipeline@v1.3.0-alltime-deepseek · 15693 in / 15758 out tokens · 140399 ms · 2026-08-03T23:19:22.105916+00:00 · methodology

0 comments
read the original abstract

Optical neural networks (ONNs) have been developed to enhance processing speed and energy efficiency in machine learning by leveraging optical devices for nonlinear activation and establishing connections among neurons. In this work, we propose a quantum optical neural network (QONN) that utilizes atom-cavity neurons with controllable photon absorption and emission. These quantum neurons are designed to replace the electronic components in ONNs, which typically introduce delays and substantial energy consumption during nonlinear activation. To evaluate the performance of the QONN, we apply it to the MNIST digit classification task, considering the effects of photon absorption duration, random atom-cavity detuning, and stochastic photon loss. Additionally, we introduce a convolutional QONN to facilitate a real-world satellite image classification (SAT-6) task. Due to its compact hardware and low power consumption, the QONN offers a promising solution for real-time satellite sensing, reducing communication bandwidth with ground stations and thereby enhancing data security.

Figures

Figures reproduced from arXiv: 2511.06167 by Chuanzhou Zhu, Daniel Soh, Peter L. McMahon, Tianyu Wang.

Figure 1
Figure 1. Figure 1: FIG. 1. Quantum optical neural network for MNIST digit classification and DeepSat (SAT-6) airborne image classification [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2. Performance of the quantum optical neural network with various photon absorption times for the MNIST task. [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: FIG. 3. Performance of the quantum optical neural network [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: FIG. 4. Quantum optical neural network with stochastic [PITH_FULL_IMAGE:figures/full_fig_p006_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: FIG. 5. Quantum optical neural network performing DeepSat (SAT-6) airborne image classification task. [PITH_FULL_IMAGE:figures/full_fig_p007_5.png] view at source ↗

discussion (0)

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

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Computational Superiority of Non-Markovian Kerr Feedback in Continuous-Variable Quantum Reservoir Computing

    quant-ph 2026-06 unverdicted novelty 6.0

    A single Kerr mode with feedback depth D achieves cross-time nonlinear rank D, exceeding the 2N hardware limit of any N-mode Gaussian reservoir.

  2. Quantum Nonlinearity for Optical Neural Computing

    physics.optics 2026-01 conditional novelty 6.0

    Embedding saturable quantum emitters in inverse-designed nanophotonic structures yields optical nonlinearity at nanowatt-per-square-micrometer intensities, enabling all-optical classification and reinforcement learnin...

Reference graph

Works this paper leans on

88 extracted references · 4 linked inside Pith · cited by 2 Pith papers

  1. [1]

    LeCun, Y

    Y. LeCun, Y. Bengio, and G. Hinton, Deep learning, Na- ture521, 436 (2015)

    2015

  2. [2]

    N. C. Thompson, K. Greenewald, K. Lee, and G. F. Manso,Thecomputationallimitsofdeeplearning(2022), arXiv:2007.05558 [cs.LG]

    Pith/arXiv arXiv 2022

  3. [3]

    N. L. Kazanskiy, M. A. Butt, and S. N. Khonina, Optical computing: Status and perspectives, Nanomaterials12, 10.3390/nano12132171 (2022)

    doi:10.3390/nano12132171 2022

  4. [4]

    Y. Shen, N. C. Harris, S. Skirlo, M. Prabhu, T. Baehr- Jones, M. Hochberg, X. Sun, S. Zhao, H. Larochelle, D.Englund,andM.Soljačić,Deeplearningwithcoherent nanophotonic circuits, Nature Photonics11, 441 (2017)

    2017

  5. [5]

    X. Lin, Y. Rivenson, N. T. Yardimci, M. Veli, Y. Luo, M. Jarrahi, and A. Ozcan, All-optical machine learning using diffractive deep neural networks, Science361, 1004 (2018)

    2018

  6. [6]

    C. Rios, N. Youngblood, Z. Cheng, M. L. Gallo, W. H. P. Pernice, C. D. Wright, A. Sebastian, and H. Bhaskaran, In-memory computing on a photonic platform, Science Advances5, eaau5759 (2019)

    2019

  7. [7]

    Wetzstein, A

    G. Wetzstein, A. Ozcan, S. Gigan, S. Fan, D. Englund, M. Soljačić, C. Denz, D. A. B. Miller, and D. Psaltis, Inference in artificial intelligence with deep optics and photonics, Nature588, 39 (2020)

    2020

  8. [8]

    X. Xu, M. Tan, B. Corcoran, J. Wu, A. Boes, T. G. Nguyen, S. T. Chu, B. E. Little, D. G. Hicks, R. Moran- dotti, A. Mitchell, and D. J. Moss, 11 tops photonic con- volutional accelerator for optical neural networks, Nature 589, 44 (2021)

    2021

  9. [9]

    Feldmann, N

    J. Feldmann, N. Youngblood, M. Karpov, H. Gehring, X. Li, M. Stappers, M. Le Gallo, X. Fu, A. Lukashchuk, A. S. Raja, J. Liu, C. D. Wright, A. Sebastian, T. J. Kippenberg, W. H. P. Pernice, and H. Bhaskaran, Paral- lel convolutional processing using an integrated photonic tensor core, Nature589, 52 (2021)

    2021

  10. [10]

    T. Zhou, X. Lin, J. Wu, Y. Chen, H. Xie, Y. Li, J. Fan, H. Wu, L. Fang, and Q. Dai, Large-scale neuromorphic optoelectronic computing with a reconfigurable diffrac- tive processing unit, Nature Photonics15, 367 (2021)

    2021

  11. [11]

    Wang, S.-Y

    T. Wang, S.-Y. Ma, L. G. Wright, T. Onodera, B. C. Richard, and P. L. McMahon, An optical neural network using less than 1 photon per multiplication, Nature Com- munications13, 123 (2022)

    2022

  12. [12]

    Ashtiani, A

    F. Ashtiani, A. J. Geers, and F. Aflatouni, An on-chip photonic deep neural network for image classification, Nature606, 501 (2022)

    2022

  13. [13]

    Sludds, S

    A. Sludds, S. Bandyopadhyay, Z. Chen, Z. Zhong, J. Cochrane, L. Bernstein, D. Bunandar, P. B. Dixon, S. A. Hamilton, M. Streshinsky, A. Novack, T. Baehr- Jones, M. Hochberg, M. Ghobadi, R. Hamerly, and D. Englund, Delocalized photonic deep learning on the internet’s edge, Science378, 270 (2022)

    2022

  14. [14]

    B. J. Shastri, A. N. Tait, T. Ferreira de Lima, W. H. P. Pernice, H. Bhaskaran, C. D. Wright, and P. R. Pruc- nal,Photonicsforartificialintelligenceandneuromorphic computing, Nature Photonics15, 102 (2021)

    2021

  15. [15]

    Hamerly, L

    R. Hamerly, L. Bernstein, A. Sludds, M. Soljačić, and D. Englund, Large-scale optical neural networks based on photoelectric multiplication, Phys. Rev. X9, 021032 (2019)

    2019

  16. [16]

    M. A. Nahmias, T. F. de Lima, A. N. Tait, H.-T. Peng, B. J. Shastri, and P. R. Prucnal, Photonic multiply- accumulateoperationsforneuralnetworks,IEEEJournal of Selected Topics in Quantum Electronics26, 1 (2020)

    2020

  17. [17]

    Bernstein, A

    L. Bernstein, A. Sludds, C. Panuski, S. Trajtenberg- Mills, R. Hamerly, and D. Englund, Single-shot optical neural network, Science Advances9, eadg7904 (2023)

    2023

  18. [18]

    Varri, F

    A. Varri, F. Brückerhoff-Plückelmann, J. Dijkstra, D. Wendland, R. Bankwitz, A. Agnihotri, and W. H. P. Pernice, Noise-resilient photonic analog neural networks, J. Lightwave Technol.42, 7969 (2024). 10

    2024

  19. [19]

    Lugnan, S

    A. Lugnan, S. Aggarwal, F. Brückerhoff-Plückelmann, C. D. Wright, W. H. P. Pernice, H. Bhaskaran, and P. Bi- enstman, Emergent self-adaptation in an integrated pho- tonic neural network for backpropagation-free learning, Advanced Science12, 2404920 (2025)

    2025

  20. [20]

    S. Choi, Y. Salamin, C. Roques-Carmes, R. Dangovski, D. Luo, Z. Chen, M. Horodynski, J. Sloan, S. Z. Uddin, and M. Soljačić, Photonic probabilistic machine learning using quantum vacuum noise, Nature Communications 15, 7760 (2024)

    2024

  21. [21]

    S.-Y. Ma, T. Wang, J. Laydevant, L. G. Wright, and P. L. McMahon, Quantum-limited stochastic optical neu- ral networks operating at a few quanta per activation, Nature Communications16, 359 (2025)

    2025

  22. [22]

    A. N. Tait, T. Ferreira de Lima, M. A. Nahmias, H. B. Miller, H.-T. Peng, B. J. Shastri, and P. R. Prucnal, Sil- icon photonic modulator neuron, Phys. Rev. Appl.11, 064043 (2019)

    2019

  23. [23]

    Stark, F

    P. Stark, F. Horst, R. Dangel, J. Weiss, and B. J. Offrein, Opportunities for integrated photonic neural networks, Nanophotonics9, 4221 (2020)

    2020

  24. [24]

    Bogaerts, D

    W. Bogaerts, D. Pérez, J. Capmany, D. A. B. Miller, J. Poon, D. Englund, F. Morichetti, and A. Melloni, Pro- grammable photonic circuits, Nature586, 207 (2020)

    2020

  25. [25]

    C. Wu, H. Yu, S. Lee, R. Peng, I. Takeuchi, and M. Li, Programmable phase-change metasurfaces on waveguides for multimode photonic convolutional neural network, Nature Communications12, 96 (2021)

    2021

  26. [26]

    Miscuglio, Z

    M. Miscuglio, Z. Hu, S. Li, J. K. George, R. Capanna, H. Dalir, P. M. Bardet, P. Gupta, and V. J. Sorger, Mas- sivelyparallelamplitude-onlyfourierneuralnetwork,Op- tica7, 1812 (2020)

    2020

  27. [27]

    J. W. Goodman, A. R. Dias, and L. M. Woody, Fully par- allel, high-speed incoherent optical method for perform- ing discrete fourier transforms, Opt. Lett.2, 1 (1978)

    1978

  28. [28]

    Psaltis, D

    D. Psaltis, D. Brady, and K. Wagner, Adaptive optical networks using photorefractive crystals, Appl. Opt.27, 1752 (1988)

    1988

  29. [29]

    J. Dong, M. Rafayelyan, F. Krzakala, and S. Gigan, Op- tical reservoir computing using multiple light scattering for chaotic systems prediction, IEEE Journal of Selected Topics in Quantum Electronics26, 1 (2020)

    2020

  30. [30]

    Chang, V

    J. Chang, V. Sitzmann, X. Dun, W. Heidrich, and G. Wetzstein, Hybrid optical-electronic convolutional neural networks with optimized diffractive optics for im- age classification, Scientific Reports8, 12324 (2018)

    2018

  31. [31]

    M. W. Matthès, P. del Hougne, J. de Rosny, G. Lerosey, and S. M. Popoff, Optical complex media as universal reconfigurable linear operators, Optica6, 465 (2019)

    2019

  32. [32]

    Larger, and D

    J.Bueno, S.Maktoobi, L.Froehly, I.Fischer, M.Jacquot, L. Larger, and D. Brunner, Reinforcement learning in a large-scale photonic recurrent neural network, Optica5, 756 (2018)

    2018

  33. [33]

    Spall, X

    J. Spall, X. Guo, T. D. Barrett, and A. I. Lvovsky, Fully reconfigurable coherent optical vector–matrix multiplica- tion, Opt. Lett.45, 5752 (2020)

    2020

  34. [34]

    Liao, M.-H

    Y. Liao, M.-H. Hsieh, and C. Ferrie, Quantum optimiza- tionfortrainingquantumneuralnetworks,QuantumMa- chine Intelligence6, 33 (2024)

    2024

  35. [35]

    J. Wen, Z. Huang, D. Cai, and L. Qian, Enhancing the expressivity of quantum neural networks with residual connections, Communications Physics7, 220 (2024)

    2024

  36. [36]

    Bischof, S

    L. Bischof, S. Teodoropol, R. M. Füchslin, and K. Stockinger, Hybrid quantum neural networks show strongly reduced need for free parameters in entity matching, Scientific Reports15, 4318 (2025)

    2025

  37. [37]

    Hirai, Practical application of quantum neural net- work to materials informatics, Scientific Reports14, 8583 (2024)

    H. Hirai, Practical application of quantum neural net- work to materials informatics, Scientific Reports14, 8583 (2024)

    2024

  38. [38]

    Thompson, Quantum neural networks can be normal, Nature Physics21, 1042 (2025)

    J. Thompson, Quantum neural networks can be normal, Nature Physics21, 1042 (2025)

    2025

  39. [39]

    García-Martín, M

    D. García-Martín, M. Larocca, and M. Cerezo, Quantum neural networks form gaussian processes, Nature Physics 21, 1153 (2025)

    2025

  40. [40]

    Chen and M

    A. Chen and M. Heyl, Empowering deep neural quantum states through efficient optimization, Nature Physics20, 1476 (2024)

    2024

  41. [41]

    I. Cong, S. Choi, and M. D. Lukin, Quantum convolu- tional neural networks, Nature Physics15, 1273 (2019)

    2019

  42. [42]

    M. S. Moreira, G. G. Guerreschi, W. Vlothuizen, J. F. Marques, J. van Straten, S. P. Premaratne, X. Zou, H. Ali, N. Muthusubramanian, C. Zachariadis, J. van Someren, M. Beekman, N. Haider, A. Bruno, C. G. Al- mudever, A. Y. Matsuura, and L. DiCarlo, Realization of a quantum neural network using repeat-until-success cir- cuits in a superconducting quantum ...

    2023

  43. [43]

    Zhang, J

    B. Zhang, J. Liu, X.-C. Wu, L. Jiang, and Q. Zhuang, Dynamical transition in controllable quantum neural net- works with large depth, Nature Communications15, 9354 (2024)

    2024

  44. [44]

    X. Pan, Z. Lu, W. Wang, Z. Hua, Y. Xu, W. Li, W. Cai, X. Li, H. Wang, Y.-P. Song, C.-L. Zou, D.-L. Deng, and L. Sun, Deep quantum neural networks on a super- conducting processor, Nature Communications14, 4006 (2023)

    2023

  45. [45]

    Zhu, Y.-D

    Y. Zhu, Y.-D. Wu, G. Bai, D.-S. Wang, Y. Wang, and G. Chiribella, Flexible learning of quantum states with generative query neural networks, Nature Communica- tions13, 6222 (2022)

    2022

  46. [46]

    K. Beer, D. Bondarenko, T. Farrelly, T. J. Osborne, R. Salzmann, D. Scheiermann, and R. Wolf, Training deep quantum neural networks, Nature Communications 11, 808 (2020)

    2020

  47. [48]

    and quantum [34] optimizers. However, quantum decoherence increases the complexity of designing reli- able quantum gates [49, 50], particularly for two-qubit gates [51, 52], which has spurred efforts to develop fault- tolerant quantum gates through topological protection [53, 54] and error correction [55]. Additionally, the need for fine-tuning internal g...

    Pith/arXiv arXiv 2025

  48. [49]

    van der Sar, Z

    T. van der Sar, Z. H. Wang, M. S. Blok, H. Bernien, T. H. Taminiau, D. M. Toyli, D. A. Lidar, D. D. Awschalom, R. Hanson, and V. V. Dobrovitski, Decoherence- protected quantum gates for a hybrid solid-state spin register, Nature484, 82 (2012)

    2012

  49. [50]

    J. R. McClean, J. Romero, R. Babbush, and A. Aspuru- Guzik, The theory of variational hybrid quantum- classical algorithms, New Journal of Physics18, 023023 (2016)

    2016

  50. [51]

    Mičuda, R

    M. Mičuda, R. Stárek, J. Fiurášek, and R. Filip, Decoherence-resilient linear optical two-qubit quantum gate, Phys. Rev. Appl.14, 054066 (2020)

    2020

  51. [52]

    X. Xue, M. Russ, N. Samkharadze, B. Undseth, A. Sam- mak, G. Scappucci, and L. M. K. Vandersypen, Quantum logic with spin qubits crossing the surface code threshold, Nature601, 343 (2022)

    2022

  52. [53]

    Kitaev, Anyons in an exactly solved model and be- yond, Annals of Physics321, 2 (2006), january Special 11 Issue

    A. Kitaev, Anyons in an exactly solved model and be- yond, Annals of Physics321, 2 (2006), january Special 11 Issue

    2006

  53. [54]

    P.Huang, N.M.Zimmerman,andG.W.Bryant,Spinde- coherence in a two-qubit cphase gate: the critical role of tunneling noise, npj Quantum Information4, 62 (2018)

    2018

  54. [55]

    Roffe, Quantum error correction: an introductory guide, Contemporary Physics60, 226 (2019)

    J. Roffe, Quantum error correction: an introductory guide, Contemporary Physics60, 226 (2019)

    2019

  55. [56]

    S. Choi, J. Kim, J. Kwak, N. Park, and S. Yu, Topologi- cally protected all-optical memory, Advanced Electronic Materials8, 2200579 (2022)

    2022

  56. [57]

    Ewaniuk, J

    J. Ewaniuk, J. Carolan, B. J. Shastri, and N. Rotenberg, Imperfect quantum photonic neural networks, Advanced Quantum Technologies6, 2200125 (2023)

    2023

  57. [58]

    G. R. Steinbrecher, J. P. Olson, D. Englund, and J. Car- olan, Quantum optical neural networks, npj Quantum Information5, 60 (2019)

    2019

  58. [59]

    Killoran, T

    N. Killoran, T. R. Bromley, J. M. Arrazola, M. Schuld, N. Quesada, and S. Lloyd, Continuous-variable quantum neural networks, Phys. Rev. Res.1, 033063 (2019)

    2019

  59. [60]

    Roncallo, A

    S. Roncallo, A. R. Morgillo, S. Lloyd, C. Macchiavello, and L. Maccone, Quantum optical shallow networks (2025), arXiv:2507.21036 [quant-ph]

    Pith/arXiv arXiv 2025

  60. [61]

    quan- tum

    M. Lewenstein and M. Olko, Storage capacity of “quan- tum” neural networks, Phys. Rev. A45, 8938 (1992)

    1992

  61. [62]

    Lewenstein and M

    M. Lewenstein and M. Olko, ’quantum’ neural networks, Network: Computation in Neural Systems2, 207 (1991)

    1991

  62. [63]

    Jennings, P

    A. Jennings, P. Horan, and J. Hegarty, Optical neural network with quantum well devices, Appl. Opt.33, 1469 (1994)

    1994

  63. [64]

    Behrman, J

    E. Behrman, J. Steck, and S. Skinner, A spatial quantum neural computer, inIJCNN’99. International Joint Conference on Neural Networks. Proceedings (Cat. No.99CH36339), Vol. 2 (1999) pp. 874–877 vol.2

    1999

  64. [65]

    Matsuda, R

    N. Matsuda, R. Shimizu, Y. Mitsumori, H. Kosaka, and K. Edamatsu, Observation of optical-fibre kerr nonlin- earity at the single-photon level, Nature Photonics3, 95 (2009)

    2009

  65. [66]

    M. Reck, A. Zeilinger, H. J. Bernstein, and P. Bertani, Experimental realization of any discrete unitary opera- tor, Phys. Rev. Lett.73, 58 (1994)

    1994

  66. [67]

    S. Choi, Y. Salamin, C. Roques-Carmes, J. Sloan, M. Horodynski, and M. Soljačić, Observing the dynam- ics of quantum states generated inside nonlinear optical cavities, Nature Communications16, 7576 (2025)

    2025

  67. [68]

    T. A. Johnson, E. Urban, T. Henage, L. Isenhower, D. D. Yavuz, T. G. Walker, and M. Saffman, Rabi oscillations between ground and rydberg states with dipole-dipole atomic interactions, Phys. Rev. Lett.100, 113003 (2008)

    2008

  68. [69]

    Yumoto, R

    G. Yumoto, R. Matsunaga, H. Hibino, and R. Shimano, Ultrafast terahertz nonlinear optics of landau level tran- sitions in a monolayer graphene, Phys. Rev. Lett.120, 107401 (2018)

    2018

  69. [70]

    Y. Lu, Y. Huang, J. Cheng, R. Ma, X. Xu, Y. Zang, Q. Wu, and J. Xu, Nonlinear optical physics at terahertz frequency, Nanophotonics13, 3279 (2024)

    2024

  70. [71]

    Stockklauser, P

    A. Stockklauser, P. Scarlino, J. V. Koski, S. Gasparinetti, C. K. Andersen, C. Reichl, W. Wegscheider, T. Ihn, K. Ensslin, and A. Wallraff, Strong coupling cavity qed with gate-defined double quantum dots enabled by a high impedance resonator, Phys. Rev. X7, 011030 (2017)

    2017

  71. [72]

    Leitenstorfer, K

    A. Leitenstorfer, K. A. Nelson, K. Reimann, and K. Tanaka, Focus on nonlinear terahertz studies, New Journal of Physics16, 045016 (2014)

    2014

  72. [73]

    Winger, A

    M. Winger, A. Badolato, K. J. Hennessy, E. L. Hu, and A. m. c. Imamoğlu, Quantum dot spectroscopy using cavity quantum electrodynamics, Phys. Rev. Lett.101, 226808 (2008)

    2008

  73. [74]

    Englund, A

    D. Englund, A. Faraon, I. Fushman, N. Stoltz, P. Petroff, and J. Vučković, Controlling cavity reflectivity with a single quantum dot, Nature450, 857 (2007)

    2007

  74. [75]

    C. Zhu, P. J. Ehlers, H. I. Nurdin, and D. Soh, Practical few-atom quantum reservoir computing, Phys. Rev. Res. 7, 023290 (2025)

    2025

  75. [76]

    Kornjaca, H.-Y

    M. Kornjaca, H.-Y. Hu, C. Zhao, J. Wurtz, P. Wein- berg, M. Hamdan, A. Zhdanov, S. H. Cantu, H. Zhou, R. A. Bravo, K. Bagnall, J. I. Basham, J. Campo, A. Choukri, R. DeAngelo, P. Frederick, D. Haines, J. Hammett, N. Hsu, M.-G. Hu, F. Huber, P. N. Jepsen, N. Jia, T. Karolyshyn, M. Kwon, J. Long, J. Lopatin, A. Lukin, T. Macri, O. Markovic, L. A. Martinez-M...

    Pith/arXiv arXiv 2024

  76. [77]

    Deng, The mnist database of handwritten digit images for machine learning research [best of the web], IEEE Signal Processing Magazine29, 141 (2012)

    L. Deng, The mnist database of handwritten digit images for machine learning research [best of the web], IEEE Signal Processing Magazine29, 141 (2012)

    2012

  77. [78]

    C. Zhu, P. J. Ehlers, H. I. Nurdin, and D. Soh, Minimal- istic and scalable quantum reservoir computing enhanced with feedback (2024), arXiv:2412.17817 [quant-ph]

    arXiv 2024

  78. [79]

    D. Soh, E. Chatterjee, and M. Eichenfield, High-fidelity state transfer between leaky quantum memories, Phys. Rev. Res.3, 033027 (2021)

    2021

  79. [80]

    S. Basu, S. Ganguly, S. Mukhopadhyay, R. DiBiano, M.Karki,andR.Nemani,Deepsat: alearningframework for satellite imagery, inProceedings of the 23rd SIGSPA- TIAL International Conference on Advances in Geo- graphic Information Systems, SIGSPATIAL 15 (Associ- ation for Computing Machinery, New York, NY, USA, 2015)

    2015

  80. [81]

    A. E. Bazarov, I. S. Goldobin, P. G. Eliseev, O. A. Ko- bilzhanov, G. T. Pak, T. V. Petrakova, T. N. Pushkina, and A. T. Semenov, Phase locking of stimulated emission fromarraysofstripegaaias/gaaslasersusingactivedirec- tional couplers, Soviet Journal of Quantum Electronics 17, 551 (1987)

    1987

Showing first 80 references.