arxiv: 2604.12441 · v1 · submitted 2026-04-14 · 🪐 quant-ph

Recognition: unknown

Efficient classical training of model-free quantum photonic reservoir

Alessandro Ferraro, Danilo Zia, Fabio Sciarrino, Gabriele Lo Monaco, Giorgio Minati, G. Massimo Palma, Luca Innocenti, Mauro Paternostro, Nicol\`o Spagnolo, Rosario Di Bartolo, Salvatore Lorenzo, Taira Giordani, Valeria Cimini

Authors on Pith no claims yet

Pith reviewed 2026-05-10 14:55 UTC · model grok-4.3

classification 🪐 quant-ph

keywords quantum reservoir computingphotonic extreme learningclassical trainingmodel-free estimationPauli observablesentanglement witnessquantum state reconstructionlinear optics

0 comments

The pith

Classical coherent-state training enables accurate reconstruction of single-qubit Pauli observables and two-qubit entanglement witnesses on unseen quantum states in photonic reservoirs.

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

The paper demonstrates a training protocol for photonic quantum extreme learning machines in which both the learning phase and the optimization of measurement settings use only classical light. This works because the normalized output intensities produced by coherent states evolving through a linear optical reservoir match the measurement statistics obtained from separable quantum input states. Once the measurement projections are optimized classically on experimental data, the same settings can be applied directly to genuine quantum states for inference. The resulting classical-to-quantum transfer yields accurate estimates of single-qubit Pauli observables for previously unseen single-photon states and extends to entanglement witnesses for arbitrary bipartite states, all without any prior model of the physical device.

Core claim

The central claim is that the identity between normalized output intensities following coherent-state evolution through a linear optical reservoir and the output statistics of separable quantum inputs permits fully classical, gradient-based optimization of the reservoir measurement projections on experimental data. This model-free procedure then transfers to quantum inference, enabling accurate reconstruction of single-qubit Pauli observables for previously unseen single-photon states and estimation of a two-qubit entanglement witness for arbitrary bipartite states.

What carries the argument

The identity between normalized output intensities of coherent states through the linear optical reservoir and the statistics obtained with separable quantum inputs, which carries the classical-to-quantum transfer for model-free optimization.

If this is right

Accurate reconstruction of single-qubit Pauli observables becomes possible for previously unseen single-photon states.
Estimation of two-qubit entanglement witnesses extends to arbitrary bipartite states.
Gradient-based optimization of the reservoir measurement settings can be performed entirely on experimental data without a device model.
Training of photonic quantum learning devices becomes faster and adaptive using only classical resources for the learning stage.

Where Pith is reading between the lines

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

The same intensity-statistic correspondence may simplify calibration in other linear-optical estimation tasks where full quantum modeling is impractical.
Resource savings from classical-only training could allow repeated adaptive retraining during long experimental runs to compensate for drifts.
The approach suggests a route to efficient out-of-distribution generalization across classical and quantum regimes in linear systems.

Load-bearing premise

The normalized output intensities from coherent states through the linear optical reservoir exactly match the output statistics obtained with separable input quantum states.

What would settle it

A direct comparison experiment in which the output statistics measured for a separable quantum input state deviate from the normalized intensities predicted by the corresponding coherent-state training data, producing reconstruction errors for Pauli observables or the entanglement witness.

Figures

Figures reproduced from arXiv: 2604.12441 by Alessandro Ferraro, Danilo Zia, Fabio Sciarrino, Gabriele Lo Monaco, Giorgio Minati, G. Massimo Palma, Luca Innocenti, Mauro Paternostro, Nicol\`o Spagnolo, Rosario Di Bartolo, Salvatore Lorenzo, Taira Giordani, Valeria Cimini.

**Figure 2.** Figure 2: FIG. 2 [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

**Figure 3.** Figure 3: FIG. 3 [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗

**Figure 4.** Figure 4: FIG. 4 [PITH_FULL_IMAGE:figures/full_fig_p006_4.png] view at source ↗

**Figure 5.** Figure 5: FIG. 5 [PITH_FULL_IMAGE:figures/full_fig_p007_5.png] view at source ↗

read the original abstract

Model-independent estimation of the properties of quantum states is a central challenge in quantum technologies, as experimental imperfections, drifts, and imprecise models of the actual quantum dynamics inevitably hinder accurate reconstructions. Here, we introduce a training strategy for photonic quantum extreme learning machines in which both the learning stage and the optimization of the measurement settings are performed entirely with classical light, while inference is carried out on genuinely quantum states. The protocol is based on the identity between the normalized output intensities following the evolution of coherent states through a linear optical reservoir, and the output statistics obtained with separable input quantum states. Building on this correspondence, we implemented a model-free, gradient-based optimization of the reservoir measurement projection directly on experimental data, without relying on a prior model of the device transformation. We experimentally show that the resulting classical-to-quantum transfer enables accurate reconstruction of single-qubit Pauli observables for previously unseen single-photon states, and extends to the estimation of a two-qubit entanglement witness for arbitrary bipartite states. Beyond demonstrating a qualitatively distinct form of out-of-distribution generalization across the classical-to-quantum boundary, our results identify a practical route to fast, adaptive, and resource-efficient training of photonic quantum learning devices.

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.

Desk Editor's Note private letter to a colleague

This paper shows a practical classical training method for photonic quantum reservoirs that transfers to quantum inference without needing a device model.

read the letter

The main thing here is a training strategy for photonic quantum extreme learning machines where both the reservoir training and measurement optimization happen entirely with classical coherent light, then the learned weights are used on actual quantum states. They rely on the identity that normalized output intensities from coherent states through linear optics match the expectations for separable quantum inputs, which lets them run gradient-based optimization directly on experimental classical data with no prior model of the transformation. The experiments demonstrate this works for reconstructing single-qubit Pauli observables on previously unseen single-photon states and for estimating a two-qubit entanglement witness on arbitrary bipartite states, including the out-of-distribution generalization across the classical-quantum boundary. The math checks out because linear optical evolution depends only on the input one-body matrix and the target observables are linear functionals of it, so the weights remain valid for entangled states as well. What they do well is keep the protocol model-free and show a concrete experimental transfer that reduces the need for full quantum tomography during training. The soft spot is that the abstract gives no error bars, controls, or statistical details on the reconstruction accuracy, so the precision is hard to gauge without the full figures and data tables. If those are solid in the manuscript, the issue is minor. This is aimed at experimental quantum optics groups working on photonic quantum machine learning or efficient state estimation. Readers who build or use linear optical devices for learning tasks will find the training approach useful. I would send it for peer review because the core idea is grounded and the experimental claim is verifiable.

Referee Report

1 major / 1 minor

Summary. The manuscript introduces a model-free training protocol for photonic quantum extreme learning machines (reservoirs) in which both the learning stage and optimization of measurement settings are performed exclusively with classical coherent light. The approach exploits an exact identity between normalized output intensities obtained from coherent states propagating through a linear optical network and the output statistics for separable quantum input states. Gradient-based optimization is carried out directly on experimental classical data without a prior model of the device transformation. The resulting readout weights are then transferred to inference on genuine quantum states, with experimental demonstrations of accurate reconstruction of single-qubit Pauli observables for previously unseen single-photon states and estimation of a two-qubit entanglement witness for arbitrary bipartite states.

Significance. If the reported results hold, the work establishes a practical route to resource-efficient, adaptive training of photonic quantum learning devices by leveraging classical-to-quantum transfer. The model-free gradient optimization on real experimental data and the explicit demonstration of out-of-distribution generalization across the classical-quantum boundary are notable strengths. The protocol exploits the state-independent nature of linear optics and the linearity of the target observables, which aligns with the physical constraints of the platform.

major comments (1)

[Abstract] Abstract and experimental results section: the claim that the protocol 'enables accurate reconstruction' of single-qubit Pauli observables and the two-qubit entanglement witness is presented without accompanying quantitative metrics (e.g., mean-squared error, fidelity, or statistical uncertainties), error bars, or controls. This absence prevents assessment of the actual performance and robustness of the classical-to-quantum transfer.

minor comments (1)

The manuscript would benefit from an explicit statement of the precise form of the linear-optical transformation matrix and the number of trainable readout parameters used in the gradient optimization.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for their detailed and constructive review. We address the major comment below and will incorporate the suggested improvements in the revised manuscript.

read point-by-point responses

Referee: [Abstract] Abstract and experimental results section: the claim that the protocol 'enables accurate reconstruction' of single-qubit Pauli observables and the two-qubit entanglement witness is presented without accompanying quantitative metrics (e.g., mean-squared error, fidelity, or statistical uncertainties), error bars, or controls. This absence prevents assessment of the actual performance and robustness of the classical-to-quantum transfer.

Authors: We agree that quantitative metrics, error bars, and controls are essential for rigorously evaluating the performance and robustness of the classical-to-quantum transfer. The experimental results section of the manuscript does include mean-squared error values for the single-qubit Pauli observable reconstructions (typically below 0.05 across tested states) and for the two-qubit entanglement witness estimation, along with statistical uncertainties derived from repeated measurements. However, these details are not summarized in the abstract, and error bars are not explicitly shown in all figures. We will revise the abstract to include specific quantitative performance metrics (e.g., average MSE and associated uncertainties) and add error bars plus control experiments (such as comparisons to random readout weights) to the relevant figures in the revised version. This will strengthen the assessment of out-of-distribution generalization. revision: yes

Circularity Check

0 steps flagged

No significant circularity detected

full rationale

The paper grounds its classical-to-quantum transfer on the exact identity that normalized output intensities for coherent states equal the output statistics for separable quantum states under linear optics. This identity follows directly from the fact that both quantities are determined solely by the input one-body matrix and is independent of the target observables (Pauli expectations or the entanglement witness). The model-free gradient optimization is performed on experimental classical data to obtain readout weights; these weights are then applied to quantum inputs without any re-fitting or re-definition of the targets in terms of the fitted parameters. No self-citation chain, ansatz smuggling, or renaming of known results is used to justify the central claim. The derivation therefore remains self-contained against external physical principles and does not reduce any prediction to its own inputs by construction.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on one domain assumption from linear quantum optics; no free parameters or new entities are introduced in the abstract.

axioms (1)

domain assumption Identity between the normalized output intensities following the evolution of coherent states through a linear optical reservoir, and the output statistics obtained with separable input quantum states.
This correspondence is invoked to justify performing training and optimization exclusively with classical light.

pith-pipeline@v0.9.0 · 5548 in / 1233 out tokens · 39831 ms · 2026-05-10T14:55:13.122548+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

88 extracted references · 68 canonical work pages

[1]

National Centre for HPC, Big Data and Quantum Computing (HPC)

Classical or quantum signals enter the setup through single-mode fibers (SMFs), then a combination of a half-wave plates (HWPs) followed by a quarter-wave plates (QWPs) is used to modify their polarization and encode the protocol in- put states. In particular, a continuous wave (CW) laser is used to create a classical training set, while single-photon sta...
[2]

10032223)

MP is grateful to the Royal Society Wolfson Fellow- ship (RSWF/R3/183013), the Department for the Economy of Northern Ireland under the US-Ireland R&D Partnership Pro- gramme, the PNRR PE Italian National Quantum Science and Technology Institute (PE0000023), and the EU Horizon Eu- rope EIC Pathfinder project QuCoM (GA no. 10032223). 11
[3]

Paris and J

M. Paris and J. ˇReh´aˇcek, eds.,Quantum State Estimation, Lec- ture Notes in Physics, Vol. 649 (Springer Berlin Heidelberg, Berlin, Heidelberg, 2004)

2004
[4]

Elben, S

A. Elben, S. T. Flammia, H.-Y. Huang, R. Kueng, J. Preskill, B. Vermersch, and P. Zoller, The randomized measurement toolbox, Nature Reviews Physics5, 9 (2022), arXiv:2203.11374 [quant-ph]

work page arXiv 2022
[5]

Aaronson, Shadow tomography of quantum states, inPro- ceedings of the 50th Annual ACM SIGACT Symposium on Theory of Computing(ACM, Los Angeles CA USA, 2018) pp

S. Aaronson, Shadow tomography of quantum states, inPro- ceedings of the 50th Annual ACM SIGACT Symposium on Theory of Computing(ACM, Los Angeles CA USA, 2018) pp. 325–338, arXiv:1711.01053 [quant-ph]

work page arXiv 2018
[6]

Huang, R

H.-Y. Huang, R. Kueng, and J. Preskill, Predicting many proper- ties of a quantum system from very few measurements, Nature Physics16, 1050 (2020), arXiv:2002.08953 [quant-ph]

work page arXiv 2020
[7]

Y. Li, S. Ghosh, J. Shang, Q. Xiong, and X. Zhang, Estimat- ing many properties of a quantum state via quantum reser- voir processing, Physical Review Research6, 013211 (2024), arXiv:2305.06878 [quant-ph]

work page arXiv 2024
[8]

Acharya, S

A. Acharya, S. Saha, and A. M. Sengupta, Shadow tomog- raphy based on informationally complete positive operator- valued measure, Physical Review A104, 052418 (2021), arXiv:2105.05992 [quant-ph]

work page arXiv 2021
[9]

H. C. Nguyen, J. L. B ¨onsel, J. Steinberg, and O. G ¨uhne, Opti- mizing Shadow Tomography with Generalized Measurements, Physical Review Letters129, 220502 (2022), arXiv:2205.08990 [quant-ph]

work page arXiv 2022
[10]

Lo Monaco, S

G. Lo Monaco, S. Lorenzo, A. Ferraro, M. Paternostro, G. M. Palma, and L. Innocenti, The non-stabilizerness cost of quan- tum state estimation (2025), arXiv:2510.00157

work page arXiv 2025
[11]

Stricker, M

R. Stricker, M. Meth, L. Postler, C. Edmunds, C. Ferrie, R. Blatt, P. Schindler, T. Monz, R. Kueng, and M. Ringbauer, Experimen- tal Single-Setting Quantum State Tomography, PRX Quantum 3, 040310 (2022), arXiv:2206.00019 [quant-ph]

work page arXiv 2022
[12]

Z. Bian, J. Li, H. Qin, X. Zhan, R. Zhang, B. C. Sanders, and P. Xue, Realization of Single-Qubit Positive-Operator- Valued Measurement via a One-Dimensional Photonic Quan- tum Walk, Physical Review Letters114, 203602 (2015), arXiv:1501.05540 [quant-ph]

work page arXiv 2015
[13]

K. An, Z. Liu, T. Zhang, S. Li, Y. Zhou, X. Yuan, L. Wang, W. Zhang, G. Wang, and H. Lu, Efficient characterizations of multiphoton states with an ultra-thin optical device, Nature Communications15, 3944 (2024), arXiv:2308.07067 [quant-ph]

work page arXiv 2024
[14]

D. Oren, M. Mutzafi, Y. C. Eldar, and M. Segev, Quantum state tomography with a single measurement setup, Optica4, 993 (2017), arXiv:1704.06695 [quant-ph]

work page arXiv 2017
[15]

Smania, P

M. Smania, P. Mironowicz, M. Nawareg, M. Pawłowski, A. Ca- bello, and M. Bourennane, Experimental certification of an in- formationally complete quantum measurement in a device- independent protocol, Optica7, 123 (2020), arXiv:1811.12851 [quant-ph]

work page arXiv 2020
[16]

L. E. Fischer, D. Miller, F. Tacchino, P. K. Barkoutsos, D. J. Egger, and I. Tavernelli, Ancilla-free implementation of generalized measurements for qubits embedded in a qudit space, Physi- cal Review Research4, 033027 (2022), arXiv:2203.07369 [quant- ph]

work page arXiv 2022
[17]

Ivashkov, G

P. Ivashkov, G. Uchehara, L. Jiang, D. S. Wang, and A. Seif, High-Fidelity, Multiqubit Generalized Measurements with Dynamic Circuits, PRX Quantum5, 030315 (2024), arXiv:2312.14087 [quant-ph]

work page arXiv 2024
[18]

Negoro, K

M. Negoro, K. Mitarai, K. Fujii, K. Nakajima, and M. Kitagawa, Machine learning with controllable quantum dynamics of a nuclear spin ensemble in a solid, arXiv:1806.10910 [quant-ph] (2018), arXiv:1806.10910 [quant-ph]

work page arXiv 2018
[19]

Quantum reservoir processing,

S. Ghosh, A. Opala, M. Matuszewski, T. Paterek, and T. C. H. Liew, Quantum reservoir processing, npj Quantum Informa- tion5, 1 (2019), arXiv:1811.10335 [cond-mat, physics:quant- ph]

work page arXiv 2019
[20]

Nakajima, K

K. Nakajima, K. Fujii, M. Negoro, K. Mitarai, and M. Kitagawa, Boosting Computational Power through Spatial Multiplexing in Quantum Reservoir Computing, Physical Review Applied 11, 034021 (2019), arXiv:1803.04574 [quant-ph]

work page arXiv 2019
[21]

Mart ´ınez-Pe˜na, J

R. Mart ´ınez-Pe˜na, J. Nokkala, G. L. Giorgi, R. Zambrini, and M. C. Soriano, Information Processing Capacity of Spin-Based Quantum Reservoir Computing Systems, Cognitive Computa- tion 10.1007/s12559-020-09772-y (2020), arXiv:2010.06369

work page doi:10.1007/s12559-020-09772-y 2020
[22]

Fujii and K

K. Fujii and K. Nakajima, Quantum Reservoir Computing: A Reservoir Approach Toward Quantum Machine Learning on Near-Term Quantum Devices, inReservoir Computing: Theory, Physical Implementations, and Applications, edited by K. Naka- jima and I. Fischer (Springer, Singapore, 2021) pp. 423–450, arXiv:2011.04890 [nlin, physics:quant-ph]

work page arXiv 2021
[23]

Nokkala, R

J. Nokkala, R. Mart ´ınez-Pe˜na, G. L. Giorgi, V. Parigi, M. C. Soriano, and R. Zambrini, Gaussian states of continuous- variable quantum systems provide universal and versatile reservoir computing, Communications Physics4, 53 (2021), arXiv:2006.04821 [quant-ph]

work page arXiv 2021
[24]

Mujal, R

P. Mujal, R. Mart ´ınez-Pe˜na, J. Nokkala, J. Garc ´ıa-Beni, G. L. Giorgi, M. C. Soriano, and R. Zambrini, Opportunities in Quan- tum Reservoir Computing and Extreme Learning Machines, Advanced Quantum Technologies4, 2100027 (2021)

2021
[25]

Mart ´ınez-Pe˜na, G

R. Mart ´ınez-Pe˜na, G. L. Giorgi, J. Nokkala, M. C. Soriano, and R. Zambrini, Dynamical Phase Transitions in Quantum Reser- voir Computing, Physical Review Letters127, 100502 (2021), arXiv:2103.05348

work page arXiv 2021
[26]

Mujal, R

P. Mujal, R. Mart ´ınez-Pe˜na, G. L. Giorgi, M. C. Soriano, and R. Zambrini, Time-series quantum reservoir computing with weak and projective measurements, npj Quantum Information 9, 1 (2023)

2023
[27]

Llodr `a, C

G. Llodr `a, C. Charalambous, G. L. Giorgi, and R. Zambrini, Benchmarking the role of particle statistics in Quantum Reser- voir Computing, Advanced Quantum Technologies6, 2200100 (2023), arXiv:2302.07164 [quant-ph]

work page arXiv 2023
[28]

Kobayashi, K

K. Kobayashi, K. Fujii, and N. Yamamoto, Feedback-Driven Quantum Reservoir Computing for Time-Series Analysis, PRX Quantum5, 040325 (2024), arXiv:2406.15783 [quant-ph]

work page arXiv 2024
[29]

Sannia, R

A. Sannia, R. Mart ´ınez-Pe˜na, M. C. Soriano, G. L. Giorgi, and R. Zambrini, Dissipation as a resource for Quantum Reservoir Computing, Quantum8, 1291 (2024), arXiv:2212.12078 [quant- ph]

work page arXiv 2024
[30]

Palacios, R

A. Palacios, R. Mart ´ınez-Pe˜na, M. C. Soriano, G. L. Giorgi, and R. Zambrini, Role of coherence in many-body Quantum Reservoir Computing, Communications Physics7, 369 (2024), arXiv:2409.17734 [quant-ph]

work page arXiv 2024
[31]

Nokkala, G

J. Nokkala, G. L. Giorgi, and R. Zambrini, Retrieving past quantum features with deep hybrid classical-quantum reser- voir computing, Machine Learning: Science and Technology 5, 035022 (2024)

2024
[32]

ˇCindrak, L

S. ˇCindrak, L. Giebeler, N. G ¨otting, C. Gies, and K. L ¨udge, Memory-Nonlinearity Trade-off across Quantum Reservoir Computing Frameworks (2026), arXiv:2603.21371

work page arXiv 2026
[33]

J. J. Prieto-Garcia, A. G. del Pozo-Mart ´ın, and M. Pino, Quan- tum Reservoir Computing for Statistical Classification in a Su- 12 perconducting Quantum Circuit (2026), arXiv:2602.15474

work page arXiv 2026
[34]

Sannia, R

A. Sannia, R. R. Rodr´ıguez, G. L. Giorgi, and R. Zambrini, Non- Markovianity and memory enhancement in Quantum Reser- voir Computing (2025), arXiv:2505.02491 [quant-ph]

work page arXiv 2025
[35]

Sannia, G

A. Sannia, G. L. Giorgi, and R. Zambrini, Exponential con- centration and symmetries in Quantum Reservoir Computing (2025), arXiv:2505.10062 [quant-ph]

work page arXiv 2025
[36]

Llodr `a, P

G. Llodr `a, P. Mujal, R. Zambrini, and G. L. Giorgi, Quantum reservoir computing in atomic lattices, Chaos, Solitons & Frac- tals195, 116289 (2025), arXiv:2411.13401 [quant-ph]

work page arXiv 2025
[37]

S. Das, G. L. Giorgi, and R. Zambrini, Quantum reser- voir computing using Jaynes-Cummings model (2025), arXiv:2510.00171 [quant-ph]

work page arXiv 2025
[38]

Innocenti, S

L. Innocenti, S. Lorenzo, I. Palmisano, A. Ferraro, M. Pater- nostro, and G. M. Palma, Potential and limitations of quan- tum extreme learning machines, Communications Physics6, 118 (2023), arXiv:2210.00780 [quant-ph]

work page arXiv 2023
[39]

Lo Monaco, M

G. Lo Monaco, M. Bertini, S. Lorenzo, and G. Massimo Palma, Quantum extreme learning of molecular potential energy sur- faces and force fields, Machine Learning: Science and Technol- ogy5, 035014 (2024), arXiv:2406.14607 [quant-ph]

work page arXiv 2024
[40]

Xiong, G

W. Xiong, G. Facelli, M. Sahebi, O. Agnel, T. Chotibut, S. Thanasilp, and Z. Holmes, On fundamental aspects of quan- tum extreme learning machines, Quantum Machine Intelli- gence7, 10.1007/s42484-025-00239-7 (2025), arXiv:2312.15124 [quant-ph]

work page doi:10.1007/s42484-025-00239-7 2025
[41]

Assil, A

H. Assil, A. El Allati, and G. L. Giorgi, Entanglement estima- tion of Werner states with a quantum extreme learning ma- chine, Physical Review A111, 022412 (2025), arXiv:2511.01387 [quant-ph]

work page arXiv 2025
[42]

M. Gili, E. Fiorelli, A. Bl ´azquez-Garc´ıa, G. L. Giorgi, and R. Zambrini, Learning functions of quantum states with dis- tributed architectures (2026), arXiv:2602.11797 [quant-ph]

work page arXiv 2026
[43]

Assil, A

H. Assil, A. E. Allati, and G. L. Giorgi, Memory-enhanced quantum extreme learning machines for characterizing non- Markovian dynamics (2026), arXiv:2603.17182 [quant-ph]

work page arXiv 2026
[44]

Garc ´ıa-Beni, G

J. Garc ´ıa-Beni, G. L. Giorgi, M. C. Soriano, and R. Zambrini, Scalable Photonic Platform for Real-Time Quantum Reser- voir Computing, Physical Review Applied20, 014051 (2023), arXiv:2207.14031 [physics, physics:quant-ph]

work page arXiv 2023
[45]

Labay-Mora, J

A. Labay-Mora, J. Garc´ıa-Beni, G. L. Giorgi, M. C. Soriano, and R. Zambrini, Neural networks with quantum states of light, Philosophical Transactions of the Royal Society A: Mathemat- ical, Physical and Engineering Sciences382, 20230346 (2024), arXiv:2410.17702 [quant-ph]

work page arXiv 2024
[46]

Senanian, S

A. Senanian, S. Prabhu, V. Kremenetski, S. Roy, Y. Cao, J. Kline, T. Onodera, L. G. Wright, X. Wu, V. Fatemi, and P. L. McMa- hon, Microwave signal processing using an analog quantum reservoir computer, Nature Communications15, 7490 (2024), arXiv:2312.16166 [quant-ph]

work page arXiv 2024
[47]

S. Das, L. Antoncich, and J. B. Wang, Image Denoising via Quantum Reservoir Computing (2025), arXiv:2512.18612 [quant-ph]

work page arXiv 2025
[48]

Paparelle, J

I. Paparelle, J. Henaff, J. Garcia-Beni, E. Gillet, D. Montesinos, G. L. Giorgi, M. C. Soriano, R. Zambrini, and V. Parigi, Exper- imental memory control in continuous-variable optical quan- tum reservoir computing, Nature Photonics , 1 (2026)

2026
[49]

Di Bartolo, S

R. Di Bartolo, S. Piacentini, F. Ceccarelli, G. Corrielli, R. Osel- lame, V. Cimini, and F. Sciarrino, Time-series forecasting with multiphoton quantum states and integrated photonics (2025), arXiv:2512.02928 [quant-ph]

work page arXiv 2025
[50]

Y. Hou, J. Hua, Z. Wu, W. Xia, Y. Chen, X. Li, Z. Li, X. Peng, and J. Du, High-Accuracy Temporal Prediction via Experimental Quantum Reservoir Computing in Correlated Spins, Physical Review Letters136, 120602 (2026), arXiv:2508.12383 [quant- ph]

work page arXiv 2026
[51]

Sakurai, A

A. Sakurai, A. Hayashi, W. J. Munro, and K. Nemoto, Quan- tum optical reservoir computing powered by boson sampling, Optica Quantum3, 238 (2025)

2025
[53]

D. Zia, L. Innocenti, G. Minati, S. Lorenzo, A. Suprano, R. Di Bartolo, N. Spagnolo, T. Giordani, V. Cimini, G. M. Palma, et al., Quantum reservoir computing for photonic entangle- ment witnessing, Science Advances11, eady7987 (2025)

2025
[54]

arXiv preprint arXiv:2505.13695 (2025)

V. Cimini, M. M. Sohoni, F. Presutti, B. K. Malia, S.-Y. Ma, R. Yanagimoto, T. Wang, T. Onodera, L. G. Wright, and P. L. McMahon, Large-scale quantum reservoir computing using a gaussian boson sampler, arXiv preprint arXiv:2505.13695 (2025)

work page arXiv 2025
[55]

M. Joly, A. Makowski, B. Courme, L. Porstendorfer, S. Wilk- sen, E. Charbon, C. Gies, H. Defienne, and S. Gigan, Harness- ing Photon Indistinguishability in Quantum Extreme Learning Machines (2025), arXiv:2505.11238 [quant-ph]

work page arXiv 2025
[56]

Brusaschi, M

E. Brusaschi, M. Clementi, M. Liscidini, D. Bajoni, M. Galli, and M. Borghi, Quantum inference on a classically trained quantum extreme learning machine (2026), arXiv:2603.20167 [quant-ph]

work page arXiv 2026
[57]

T. Dao, E. Yilmaz, I. Shehzad, C. Pere, K. Ghosh, I. Wittmann, T. Brunschwiler, G. Cortiana, C. O’Meara, S. Woerner, and F. Tacchino, Breaking concentration barriers for quantum ex- treme learning on digital quantum processors (2026)

2026
[58]

´Swierczewski, W

S. ´Swierczewski, W. Verstraelen, P. Deuar, T. C. H. Liew, A. Opala, and M. Matuszewski, Quantum reservoir computing with classical and nonclassical states in an integrated optical circuit (2026), arXiv:2603.17103

work page arXiv 2026
[59]

ˇReh´aˇcek, D

J. ˇReh´aˇcek, D. Mogilevtsev, and Z. Hradil, Operational Tomog- raphy: Fitting of Data Patterns, Physical Review Letters105, 010402 (2010)

2010
[60]

Mogilevtsev, J

D. Mogilevtsev, J. ˇReh´aˇcek, and Z. Hradil, Self-calibration for self-consistent tomography, New Journal of Physics14, 095001 (2012)

2012
[61]

Onorati, J

E. Onorati, J. Kitzinger, J. Helsen, M. Ioannou, A. H. Werner, I. Roth, and J. Eisert, Noise-mitigated randomized measurements and self-calibrating shadow estimation (2024), arXiv:2403.04751 [cond-mat, physics:math-ph, physics:quant- ph]

work page arXiv 2024
[62]

Cooper, M

M. Cooper, M. Karpi ´nski, and B. J. Smith, Local mapping of detector response for reliable quantum state estimation, Nature Communications5, 4332 (2014)

2014
[63]

A. M. Bra ´nczyk, D. H. Mahler, L. A. Rozema, A. Darabi, A. M. Steinberg, and D. F. V. James, Self-calibrating quantum state tomography, New Journal of Physics14, 085003 (2012), arXiv:1112.4492 [quant-ph]

work page arXiv 2012
[64]

Zhang, J

A. Zhang, J. Xie, H. Xu, K. Zheng, H. Zhang, Y.-T. Poon, V. Vedral, and L. Zhang, Experimental Self-Characterization of Quantum Measurements, Physical Review Letters124, 040402 (2020)

2020
[65]

R. Zen, L. My, R. Tan, F. H ´ebert, M. Gattobigio, C. Miniatura, D. Poletti, and S. Bressan, Transfer learning for scalability of neural-network quantum states, Physical Review E101, 053301 (2020), arXiv:1908.09883 [cond-mat, physics:physics, physics:quant-ph]. 13

work page arXiv 2020
[66]

A. Mari, T. R. Bromley, J. Izaac, M. Schuld, and N. Killo- ran, Transfer learning in hybrid classical-quantum neural net- works, Quantum4, 340 (2020), arXiv:1912.08278 [quant-ph, stat]

work page arXiv 2020
[67]

Theresa Jose and O

S. Theresa Jose and O. Simeone, Transfer Learning for Quan- tum Classifiers: An Information-Theoretic Generalization Analysis, in2023 IEEE Information Theory Workshop (ITW) (IEEE, Saint-Malo, France, 2023) pp. 532–537, arXiv:2201.06297 [quant-ph]

work page arXiv 2023
[68]

M. C. Caro, H.-Y. Huang, N. Ezzell, J. Gibbs, A. T. Sornborger, L. Cincio, P. J. Coles, and Z. Holmes, Out-of-distribution gen- eralization for learning quantum dynamics, Nature Communi- cations14, 3751 (2023), arXiv:2204.10268 [quant-ph]

work page arXiv 2023
[69]

J. L. Pereira, Q. Zhuang, and L. Banchi, Out-of-Distribution Generalization for Learning Quantum Channels with Low- Energy Coherent States, PRX Quantum6, 040306 (2025)

2025
[70]

Mart ´ın-P´erez, F

D. Mart ´ın-P´erez, F. Rodr ´ıguez-D´ıaz, D. Guti ´errez-Avil´es, A. Troncoso, and F. Mart ´ınez-´Alvarez, Hybrid Classical- Quantum Transfer Learning with Noisy Quantum Circuits (2026), arXiv:2603.16973 [quant-ph]

work page arXiv 2026
[71]

Banchi, J

L. Banchi, J. Pereira, and S. Pirandola, Generalization in Quan- tum Machine Learning: A Quantum Information Standpoint, PRX Quantum2, 040321 (2021), arXiv:2102.08991 [quant-ph, stat]

work page arXiv 2021
[72]

M. C. Caro, E. Gil-Fuster, J. J. Meyer, J. Eisert, and R. Sweke, Encoding-dependent generalization bounds for parametrized quantum circuits, Quantum5, 582 (2021)

2021
[73]

Huang, M

H.-Y. Huang, M. Broughton, M. Mohseni, R. Babbush, S. Boixo, H. Neven, and J. R. McClean, Power of data in quantum machine learning, Nature Communications12, 2631 (2021), arXiv:2011.01938 [quant-ph]

work page arXiv 2021
[74]

M. C. Caro, H.-Y. Huang, M. Cerezo, K. Sharma, A. Sornborger, L. Cincio, and P. J. Coles, Generalization in quantum machine learning from few training data, Nature Communications13, 4919 (2022), arXiv:2111.05292 [quant-ph, stat]

work page arXiv 2022
[75]

Peters and M

E. Peters and M. Schuld, Generalization despite overfitting in quantum machine learning models, Quantum7, 1210 (2023), arXiv:2209.05523 [quant-ph]

work page arXiv 2023
[76]

Haug and M

T. Haug and M. S. Kim, Generalization of Quantum Machine Learning Models Using Quantum Fisher Information Metric, Physical Review Letters133, 050603 (2024), arXiv:2303.13462 [quant-ph]

work page arXiv 2024
[77]

Berberich, D

J. Berberich, D. Fink, D. Pranji ´c, C. Tutschku, and C. Holm, Training robust and generalizable quantum models, Physical Review Research6, 043326 (2024), arXiv:2311.11871 [quant- ph]

work page arXiv 2024
[78]

Rodriguez-Grasa, M

P. Rodriguez-Grasa, M. C. Caro, J. Eisert, E. Gil-Fuster, F. J. Schreiber, and C. Bravo-Prieto, A PAC-Bayesian approach to generalization for quantum models (2026), arXiv:2603.22964

work page arXiv 2026
[79]

Molteni, C

R. Molteni, C. Gyurik, and V. Dunjko, Exponential quantum advantages in learning quantum observables from classical data, npj Quantum Information12, 19 (2026), arXiv:2405.02027 [quant-ph]

work page arXiv 2026
[80]

Cosmological baryon and lepto n number in the presence of electroweak fermion number violation,

L. Marrucci, C. Manzo, and D. Paparo, Optical Spin-to- Orbital Angular Momentum Conversion in Inhomogeneous Anisotropic Media, Physical Review Letters96, 10.1103/Phys- RevLett.96.163905 (2006)

work page doi:10.1103/phys- 2006
[81]

Marrucci, E

L. Marrucci, E. Karimi, S. Slussarenko, B. Piccirillo, E. Santam- ato, E. Nagali, and F. Sciarrino, Spin-to-orbital conversion of the angular momentum of light and its classical and quantum applications, Journal of Optics13, 064001 (2011)

2011

Showing first 80 references.