Enhanced image classification via hybridizing quantum dynamics with classical neural networks

Abolfazl Bayat; Ruiyang Zhou; Saubhik Sarkar; Sougato Bose

REVIEW 3 major objections 6 minor 1 cited by

Enhanced image classification via hybridizing quantum dynamics with classical neural networks

T0 review · 3 major / 6 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read This paper argues that time-evolving encoded images under a tiny Ising spin chain makes classes nearly orthogonal in Hilbert space, and that this quantum dynamics—not the classical encoder alone—is responsible for the accuracy gains on…

desk verdict The hybrid protocol is a plausible heuristic, but the central claim that the quantum module drives the accuracy gain rests on an underpowered classical baseline, so the paper needs major revision before it can be accepted. read the letter →

arxiv 2507.13587 v1 pith:M2KWU5EE submitted 2025-07-18 quant-ph cond-mat.dis-nn

Ruiyang Zhou , Saubhik Sarkar , Sougato Bose , Abolfazl Bayat This is my paper

classification quant-phcond-mat.dis-nn PACS 03.67.Lx

keywords quantummachinelearningimageclassificationtransverse-fieldIsingmodelnon-equilibriumdynamicshybridclassical-quantumprotocolfidelityshadowtomography

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper proposes a hybrid architecture in which a small classical network first compresses an image to a few numbers, and those numbers set the magnetic fields of a short transverse-field Ising chain whose time evolution becomes the classifier. The claim is that the quantum evolution maps images from different classes to nearly orthogonal quantum states, so class membership can be read out by comparing a test state with per-class averaged projectors. On SAT6, BloodMNIST, MNIST, and Fashion-MNIST, the hybrid is reported to reach 92.38%, 74.97%, 95.63%, and 83.86% test accuracy respectively, while the classical encoder alone reaches 85.40%, 50.27%, 65.43%, and 68.78%. If correct, this is a parameter-free use of many-body dynamics that bypasses variational training and points toward practical quantum advantage for near-term devices.

What carries the argument

The workhorse is a transverse-field Ising chain, $H = J \sum_j \sigma^x_j \sigma^x_{j+1} + \sum_j h_j \sigma^z_j$ with $J=1$, where the site-dependent fields $h = (h_1,\dots,h_{N_q})$ are the output of a single-hidden-layer classical encoder. The initial state $|0,\dots,0\rangle$ evolves for time $t$; fidelity between two evolved states is computed by swap test or direct swap-operator measurement, and a quantum comparator trains the encoder by minimizing $|F - \delta_{y_1,y_2}|$ over image pairs. Classification then uses per-class observables $O_\alpha = \frac{1}{N_S}\sum_k |\Psi_k^{(\alpha)}(t)\rangle\langle\Psi_k^{(\alpha)}(t)|$, assigning the label whose averaged expectation on the test state is largest.

What would settle it

On MNIST with 50,000 training samples, train a classical network with one hidden layer of 256 neurons, ten output neurons, and a cross-entropy loss under the same training schedule; if its test accuracy equals or exceeds the hybrid's 95.63%, the claim that the quantum module is decisive is contradicted. A complementary check is to measure the average fidelity between different-class states at evolution time $t = 2N_q/J$; if it is not close to zero, the orthogonality mechanism described in the paper is not actually in operation.

Watch

Extended reading notes

Core claim

The central claim is that non-equilibrium quantum dynamics is itself a discriminative resource for classical image data. Starting from an identical product state, two systems evolved under slightly different magnetic-field configurations lose fidelity rapidly, so encoding different image classes into different fields yields almost-orthogonal states that a single measurement can distinguish. The paper reports that after training only a classical encoder to make fidelity match class equality, the full hybrid classifier outperforms the classical module on every dataset tested, with the largest gap on MNIST (95.63% versus 65.43%). The authors state explicitly that the quantum module plays the key role in this enhancement beyond what the classical module can achieve.

Load-bearing premise

The load-bearing premise is that the classical network used for comparison—the same encoder architecture trained with mean-squared error and as many outputs as qubits—sets the right performance ceiling, so if a standard classification-trained network of similar size and loss could match the hybrid's accuracy, the paper's central claim of quantum-caused improvement would fail.

Editorial extensions

If this is right

With only three or four qubits, the hybrid classifier reaches 92.38% on SAT6, 74.97% on BloodMNIST, 95.63% on MNIST, and 83.86% on Fashion-MNIST, all above the classical module alone.
The quantum module has no trainable parameters, so the classical encoder is the only part optimized during training, and the protocol avoids gradient-based tuning of quantum gates.
High accuracy appears once the evolution time exceeds about $N_q/4J$ and the Hilbert-space dimension satisfies $2^{N_q} \ge N_c$, consistent with the orthogonality mechanism the paper proposes.
Shadow tomography can replace exact fidelity measurements with modest sample counts, and on simpler datasets it closely matches the exact-measurement accuracy of the swap-test protocol.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

A consequence the paper leaves implicit is that, because training touches only the classical encoder, barren-plateau arguments that apply to variational quantum circuits are not directly relevant; whether the scheme keeps its edge under hardware noise is an open question.
I infer the protocol is effectively a data-dependent quantum kernel: classification is decided by fidelities between a test state and per-class training states, so a direct comparison with classical kernel methods on the same encoder outputs would isolate what the Hilbert-space map adds.
A testable extension would be to increase $N_q$ and the number of classes while checking that accuracy continues to track the conditions $2^{N_q} \ge N_c$ and $t \gtrsim N_q/4J$; a premature saturation of accuracy would mean the orthogonality picture alone is not the explanation.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note

read the letter

The headline result here doesn't survive contact with the baseline. The paper reports that adding quantum dynamics to a classical encoder boosts MNIST accuracy from 65% to 95%, and uses that gap to claim the quantum module is key. But the classical comparison in Fig. 6 is the same encoder with Nq output neurons trained with MSE. For MNIST that's four outputs trying to separate ten classes—it isn't a reasonable classifier. So the gap mostly tells you the baseline is underpowered, not that the quantum dynamics matter.

That said, the paper is a serious empirical proposal. The architecture is a trainable classical encoder with a pairwise fidelity comparator loss, followed by fixed transverse-field Ising dynamics and class-averaged observables. I don't see that exact combination in the existing quantum reservoir or kernel literature. The numerical study is also careful: four datasets, SEM over 10 runs, and systematic sweeps over evolution time, qubit number, sample size, hidden layers, and training set size, with both swap test and shadow tomography variants. For a heuristic QML paper, that's a solid level of rigor.

The soft spots beyond the baseline are important. For the qubit numbers used (3–4), the whole feature map is classically simulable at trivial cost. The authors never test a classical kernel or random-feature model built on the same h features, which would almost certainly reproduce the hybrid results. So the phrase 'practical quantum advantage' in the abstract is not supported by anything in the paper. Also, no code or data are released, so the specific numbers can't be checked.

Is the protocol interesting anyway? Yes, as a quantum reservoir-style feature map with a comparator-trained encoder. Readers working on quantum kernel methods or reservoir computing will want to see this. But the central claim needs to be rebuilt around a realistic classical baseline: a version of the encoder with Nc outputs and cross-entropy loss, and ideally a classical feature map with comparable nonlinear structure. That's a major revision, not a polish.

I'd send it to peer review—it's a legitimate contribution that deserves referee time—but I wouldn't accept it in its current form. The authors should also soften the advantage claim, release code, and add the missing classical baselines.

Referee Report

3 major / 6 minor

Summary. The paper proposes a hybrid classical-quantum classifier for image data. A fully connected encoder maps each image to Nq magnetic-field parameters h; starting from the product state |0...0>, the system evolves under the transverse-field Ising Hamiltonian H(h) for time t. During training, pairs of images are compared by computing the fidelity of their evolved states against the label-equality indicator (Eq. 2). At inference, class observables O_alpha are formed by averaging projectors over NS training states per class (Eq. 3), and a test image is assigned the class whose observable has the largest expectation value. Experiments on SAT6, BloodMNIST, MNIST, and Fashion-MNIST report accuracies of 92.38%, 74.97%, 95.63%, and 83.86% using 3-4 qubits. Section VI compares the hybrid model against the classical encoder alone and concludes that the quantum module is essential and that the protocol has potential for practical quantum advantage.

Significance. If the reported accuracies were accompanied by a faithful classical reference, this would be a useful contribution to quantum machine learning: the architecture cleanly combines a classical encoder with a physically motivated nonlinear quantum feature map, and the paper presents a careful comparison of swap-test and shadow-tomography readouts. The numerical experiments are described clearly and include standard errors over 10 runs, together with ablation sweeps over NS, t, Nq, hidden-layer count, and training-set size, which is a strength. However, the central comparative claim is not currently established: the baseline in Sec. VI is not a standard classifier, and for the Nq<=4 systems used here the quantum feature map is exactly classically simulable. The paper's significance as a demonstration of quantum advantage therefore hinges on additional comparisons that are not yet present.

major comments (3)

[Sec. VI, Fig. 6] The comparison intended to show the "key role" of the quantum module uses a classical baseline that is not a well-specified classifier. As stated in Sec. VI, the baseline keeps the same geometry and "only change[s] the loss function to the standard mean squared error," so it has Nq output neurons (3 or 4) rather than Nc outputs (6, 8, or 10), and for every dataset Nq<Nc. Such a network cannot represent one-hot class labels, and the manuscript never specifies the regression targets used for this MSE training. The large gaps in Fig. 6 (e.g., 65.43% vs 95.63% for MNIST) can therefore be explained by an underpowered and mis-specified classical reference rather than by the quantum dynamics. Please replace this baseline with the same encoder trained with cross-entropy and Nc outputs, and additionally report standard classical classifiers (logistic regression, SVM, multi-layer perceptron) and classical kernel or random-feature models built from the same h-feature map.
[Sec. V.C / Table I] The quantum module used in the headline results is a 3- or 4-qubit transverse-field Ising evolution, which is exactly and efficiently simulable on a classical computer. Since all reported experiments are numerical simulations, the protocol's decision rule is, in effect, a classical nonlinear map h -> {fidelity with class-averaged states}; the simulation cost scales as 2^Nq and is trivial for Nq<=4. Consequently, the accuracy gap in Fig. 6 does not by itself indicate any computational advantage of quantum hardware. The authors should either (i) benchmark classical kernel, Random Fourier feature, or SVM models using the same Nq-dimensional h-features, or (ii) substantially temper the "practical quantum advantage" language in the abstract and conclusion to claims about the accuracy of a specific hybrid architecture.
[Sec. V.B / Sec. V.C / Table I] The paper does not describe a validation split or hyperparameter selection procedure. The sweeps over evolution time t (Fig. 5a) and qubit number Nq (Fig. 5b), together with fixed choices NS=100, hidden width 256, and tJ=2Nq, appear to be informed by test-set accuracy, and Table I then reports test accuracy for those choices. If the test set was used for model selection, the reported accuracies are optimistically biased, and the SEM over 10 runs does not correct for this. Please state clearly whether a validation set was used for selecting Nq, t, and NS, or provide a nested train/validation/test protocol.

minor comments (6)

[Sec. III] The first paragraph contains a typo: "precious section" should be "previous section."
[Fig. 5(b)] Each series label in the legend includes "Nq=8" even though the horizontal axis is Nq; the legend should instead identify the dataset, since the panel sweeps Nq.
[Eq. (5)] After expanding the controlled-swap state in the |±> basis, the coefficients are 1/2 rather than 1/sqrt2; the resulting probabilities p± are as stated, but the expression as written is not normalized.
[Sec. IV] MNIST is cited as [112] and Fashion-MNIST as [113], but reference [112] is the Fashion-MNIST paper and [113] is the MNIST paper; the citations are swapped.
[Fig. S1] The text says "the architecture begins with a fixed input layer (mapped from the quantum measurement outcomes)," which conflicts with the main protocol where the classical module receives raw images; please clarify whether this ablation adds post-quantum layers.
[Reproducibility] No code or data availability statement is included; providing the simulation code would strengthen reproducibility given the many hyperparameters and the numerical nature of the results.

Circularity Check

2 steps flagged · score 4.0 of 10

The claimed key role of the quantum module is forced by a classical baseline with Nq<Nc outputs and MSE loss, so the accuracy gap is a comparison-design artifact.

other [Section VI, Fig. 6; Section III A (classical module output dimension)]
"As a first check, we can solely use our classical neural network for solving the classification task. This shows the maximum capacity of the classical neural network that is used in our protocol for classifying the images. In order to have a fair comparison, we use the same number of neurons with the same geometry and only change the loss function to the standard mean squared error. ... The number of neurons in the output layer is equal to the number of qubits Nq in the quantum Ising chain."

The 'classical module alone' baseline keeps the hybrid encoder's Nq-dimensional output layer and switches to MSE loss. For all datasets Nq<Nc (SAT6: 3<6, BloodMNIST: 3<8, MNIST and Fashion-MNIST: 4<10), so the baseline cannot represent the Nc class labels under a standard regression target. Its low accuracy is therefore a structural consequence of the baseline definition. The paper uses that low accuracy to conclude that 'the quantum module plays a key role in enhancing the classification accuracy beyond the capabilities of the classical module alone.' The claimed enhancement is fixed by the underpowered comparison, not independently demonstrated by the quantum dynamics.
other [Section VI, Fig. 7]
"Further confirmation of the advantage of using the quantum module can be seen by performing principal component analysis (PCA) of the h-fields from the classical module and the measurement outputs of the full hybrid model. ... while the outputs of the classical module are not quite distinguished by the PCA, the quantum module is able to classify the data in a lot more convincing manner."

The PCA compares raw h-fields (Nq-dimensional intermediate encodings) with the protocol's final Nc-dimensional measurement score vector. Those scores are exactly what the trained comparator and observables were optimized to separate, so a PCA plot on the scores cannot serve as independent evidence that the quantum dynamics, rather than the training objective, is responsible for the separation. This is a second instance of the same mismatched-comparison design.

full rationale

The paper's training and classification pipeline is self-contained: the quantum comparator loss (Eq. 2), the fidelity-based observables (Eq. 3), and the test-accuracy evaluation (Eq. 4) constitute a legitimate supervised learning scheme, and no analytical result reduces to a fitted value by equation. No load-bearing self-citation or imported uniqueness theorem appears; self-citations to Banchi-Bayat-Bose concern measurement techniques and are not the basis of the central claim. The circularity lies in the Section VI demonstration: the classical baseline is deliberately the same Nq-output encoder retrained with MSE, and because Nq<Nc for every dataset it cannot encode all class labels, so its poor accuracy is guaranteed. The PCA 'confirmation' repeats the mismatch by comparing intermediate h-fields to final optimized score vectors. The central claim that the quantum module is essential is therefore supported only by a comparison that forces the conclusion, corresponding to a partial circularity score of 4 rather than a fully fabricated derivation.

Assumptions & free parameters 5 free parameters · 5 assumptions · 0 invented entities

The protocol depends on standard quantum mechanics, the transverse Ising Hamiltonian, and the chosen loss and observables. It introduces no new physical entities. The main hand-chosen quantities are the evolution time, qubit number, hidden layer width, sample counts, and trained network weights; several are selected after scanning test accuracy rather than on a validation set.

free parameters (5)

Encoder network weights and biases = not reported
Trained on 50,000 image pairs per dataset via the fidelity loss in Eq. (2). The central accuracy numbers depend on these learned parameters.
Evolution time t = tJ = 2 N_q
Selected after scanning test accuracy versus time in Fig. 5a; no validation-based selection is reported.
Number of qubits N_q = 3 for SAT6/BloodMNIST, 4 for MNIST/Fashion-MNIST
Chosen to satisfy 2^N_q >= N_c and confirmed by the test-accuracy scan in Fig. 5b.
Hidden layer width = 256
Fixed by hand; only depth (0 to 4 layers) is studied in Fig. S1, not width.
Measurement samples N_S and shadow shots M = N_S = 100, M = 5000
Chosen to exceed convergence thresholds observed in Fig. 4; these affect the reported numbers.

assumptions (5)

standard math Standard postulates of quantum mechanics, unitary evolution under Eq. (1), and the swap test and shadow tomography formulas.
The protocol assumes these standard results without proof.
domain assumption The classical encoder maps images to real magnetic field values that can be implemented as site-dependent fields in an analog Ising simulator.
Section III assumes the h-vector from the neural network is physically realizable with sufficient precision; no hardware demonstration is given.
domain assumption The fidelity between states evolved under different field configurations decays as shown in Fig. 1, and the training can reach a low-loss regime that generalizes to test images.
Section II demonstrates the decay for random fields, but the trainability and generalization of the full encoder is not proven.
domain assumption Noiseless quantum evolution is a valid proxy for NISQ hardware performance.
All results come from exact simulations; no noise model or error mitigation is included.
ad hoc to paper The observables O_alpha in Eq. (3), constructed as uniform averages over N_S training states, are a good choice for classification.
This is a design choice; no optimality argument is given.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Enhanced image classification via hybridizing quantum dynamics with classical neural networks." pith.science (2026). https://pith.science/paper/M2KWU5EE

@misc{pith2026250713587,
  author       = {Pith},
  title        = {Pith review of: Enhanced image classification via hybridizing quantum dynamics with classical neural networks},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/M2KWU5EE}},
  note         = {Machine review of arXiv:2507.13587}
}

read the original abstract

The integration of quantum computing and machine learning has emerged as a promising frontier in computational science. We present a hybrid protocol which combines classical neural networks with non-equilibrium dynamics of a quantum many-body system for image classification. This architecture leverages classical neural networks to efficiently process high-dimensional data and encode it effectively on a quantum many-body system, overcoming a challenging task towards scaled up quantum computation. The quantum module further capitalizes on the discriminative properties of many-body quantum dynamics to enhance classification accuracy. By mapping images from distinct classes to nearly-orthogonal quantum states, the system maximizes separability in the Hilbert space, enabling robust classification. We evaluate the performance of our model on several benchmark datasets with various number of features and classes. Moreover, we demonstrate the key role of the quantum module in achieving high classification accuracy which cannot be accomplished by the classical neural network alone. This showcases the potential of our hybrid protocol for achieving practical quantum advantage and paves the way for future advancements in quantum-enhanced computational techniques.

Figures

Figures reproduced from arXiv: 2507.13587 by the authors.

**Figure 2.** FIG. 2. Protocol by the hybrid classical-quantum neural network. (a) The quantum comparator. (b) The quantum classifier. Both the quantum [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

**Figure 3.** FIG. 3. Data samples from the (a) SAT6, (b) BloodMNIST, (c) MNIST, and (d) Fashion-MNIST datasets. (a) SAT6, containing six distinct [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗

Figures from the paper (4 more)

**Figure 4.** Figure 4: FIG. 4. Impact of the number of samples to construct the observables on the classification accuracy for the (a) SAT6, (b) BloodMNIST, (c) [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]

**Figure 5.** Figure 5: FIG. 5. The dependence of the classification accuracy for the SAT6, BloodMNIST, MNIST, and Fashion-MNIST datasets on (a) the evolution [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]

**Figure 6.** Figure 6: FIG. 6. The classification accuracy of only the classical module and [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]

**Figure 7.** Figure 7: FIG. 7. Distinguishability comparison by the PCA on the SAT6, BloodMNIST, MNIST, and Fashion-MNIST datasets with [PITH_FULL_IMAGE:figures/full_fig_p009_7.png]

Discussion (0). Sign in to comment.

Forward citations

Cited by 1 Pith paper

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

Quantum feature-map learning with reduced resource overhead
quant-ph 2025-10 conditional novelty 6.0 of 10

By classically reconstructing quantum model outputs, Q-FLAIR selects gates, features, and weights with O(M) quantum evaluations per iteration, decoupling quantum cost from feature dimension and enabling >90% MNIST acc...

Reference graph

Works this paper leans on

117 extracted references · 71 canonical work pages · cited by 1 Pith paper

[1]

Algorithms for quantum computation: discrete logarithms and factoring,

P. W. Shor, “Algorithms for quantum computation: discrete logarithms and factoring,” in Proceedings 35th annual sym- posium on foundations of computer science (IEEE, 1994) pp. 124–134

1994
[2]

A fast quantum mechanical algorithm for database search,

L. K. Grover, “A fast quantum mechanical algorithm for database search,” in Proceedings of the twenty-eighth annual ACM symposium on Theory of computing (1996) pp. 212–219

1996
[3]

Quantum algo- rithm for linear systems of equations,

A. W. Harrow, A. Hassidim, and S. Lloyd, “Quantum algo- rithm for linear systems of equations,” Phys. Rev. Lett. 103, 150502 (2009)

2009
[4]

Stabilizer Codes and Quantum Error Correc- tion,

D. Gottesman, “Stabilizer Codes and Quantum Error Correc- tion,” arXiv:9705052 (1997)

1997
[5]

Theory of quantum error- correcting codes,

E. Knill and R. Laflamme, “Theory of quantum error- correcting codes,” Physical Review A55, 900 (1997)

1997
[6]

Fault-tolerant quantum computation by anyons,

A. Y . Kitaev, “Fault-tolerant quantum computation by anyons,” Annals of physics 303, 2–30 (2003)

2003
[7]

Noisy intermediate-scale quantum algo- rithms,

K. Bharti, A. Cervera-Lierta, T. H. Kyaw, T. Haug, S. Alperin- Lea, A. Anand, M. Degroote, H. Heimonen, J. S. Kottmann, T. Menke, et al. , “Noisy intermediate-scale quantum algo- rithms,” Reviews of Modern Physics94, 015004 (2022)

2022
[8]

Quantum computing in the NISQ era and beyond,

J. Preskill, “Quantum computing in the NISQ era and beyond,” Quantum 2, 79 (2018)

2018

Show all 117 references

[9]

Progress towards practical quantum variational algorithms,

D. Wecker, M. B. Hastings, and M. Troyer, “Progress towards practical quantum variational algorithms,” Physical Review A 92, 042303 (2015)

2015
[10]

The theory of variational hybrid quantum-classical al- gorithms,

J. R. McClean, J. Romero, R. Babbush, and A. Aspuru- Guzik, “The theory of variational hybrid quantum-classical al- gorithms,” New Journal of Physics 18, 023023 (2016)

2016
[11]

Variational quantum algorithms,

M. Cerezo, A. Arrasmith, R. Babbush, S. C. Benjamin, S. Endo, K. Fujii, J. R. McClean, K. Mitarai, X. Yuan, L. Cin- cio, et al., “Variational quantum algorithms,” Nature Reviews 10 Physics 3, 625–644 (2021)

2021
[12]

A vari- ational eigenvalue solver on a photonic quantum processor,

A. Peruzzo, J. McClean, P. Shadbolt, M.-H. Yung, X.-Q. Zhou, P. J. Love, A. Aspuru-Guzik, and J. L. O’brien, “A vari- ational eigenvalue solver on a photonic quantum processor,” Nat. Commun. 5, 4213 (2014)

2014
[13]

Vari- ational quantum state eigensolver,

M. Cerezo, K. Sharma, A. Arrasmith, and P. J. Coles, “Vari- ational quantum state eigensolver,” npj Quantum Information 8, 113 (2022)

2022
[14]

Symmetry en- hanced variational quantum spin eigensolver,

C. Lyu, X. Xu, M.-H. Yung, and A. Bayat, “Symmetry en- hanced variational quantum spin eigensolver,” Quantum 7, 899 (2023)

2023
[15]

Multilevel variational spectroscopy using a programmable quantum sim- ulator,

Z. Han, C. Lyu, Y . Zhou, J. Yuan, J. Chu, W. Nuerbolati, H. Jia, L. Nie, W. Wei, Z. Yang, L. Zhang, Z. Zhang, C.-K. Hu, L. Hu, J. Li, D. Tan, A. Bayat, S. Liu, F. Yan, and D. Yu, “Multilevel variational spectroscopy using a programmable quantum sim- ulator,” Phys. Rev. Res.6,...

2024
[16]

Scaling of variational quantum circuit depth for condensed matter systems,

C. Bravo-Prieto, J. Lumbreras-Zarapico, L. Tagliacozzo, and J. I. Latorre, “Scaling of variational quantum circuit depth for condensed matter systems,” Quantum 4, 272 (2020)

2020
[17]

Variational quan- tum eigensolver for frustrated quantum systems,

A. Uvarov, J. D. Biamonte, and D. Yudin, “Variational quan- tum eigensolver for frustrated quantum systems,” Phys. Rev. B 102, 075104 (2020)

2020
[18]

Classically optimized variational quantum eigensolver with applications to topological phases,

K. N. Okada, K. Osaki, K. Mitarai, and K. Fujii, “Classically optimized variational quantum eigensolver with applications to topological phases,” Physical Review Research5 (2023)

2023
[19]

Ground-state energy estimation of the water molecule on a trapped-ion quantum computer,

Y . Nam, J.-S. Chen, N. C. Pisenti, K. Wright, C. Delaney, D. Maslov, K. R. Brown, S. Allen, J. M. Amini, J. Apisdorf, et al., “Ground-state energy estimation of the water molecule on a trapped-ion quantum computer,” npj Quantum Informa- tion 6, 33 (2020)

2020
[20]

Hardware-e fficient varia- tional quantum eigensolver for small molecules and quantum magnets,

A. Kandala, A. Mezzacapo, K. Temme, M. Takita, M. Brink, J. M. Chow, and J. M. Gambetta, “Hardware-e fficient varia- tional quantum eigensolver for small molecules and quantum magnets,” Nature 549, 242–246 (2017)

2017
[21]

Hartree-fock on a su- perconducting qubit quantum computer,

G. A. Quantum, Collaborators*†, F. Arute, K. Arya, R. Bab- bush, D. Bacon, J. C. Bardin, R. Barends, S. Boixo, M. Broughton, B. B. Buckley, et al. , “Hartree-fock on a su- perconducting qubit quantum computer,” Science 369, 1084– 1089 (2020)

2020
[22]

A quantum approx- imate optimization algorithm,

E. Farhi, J. Goldstone, and S. Gutmann, “A quantum approx- imate optimization algorithm,” arXiv:1411.4028 (2014)

2014 arXiv
[23]

Ising formulations of many NP problems,

A. Lucas, “Ising formulations of many NP problems,” Fron- tiers in physics 2, 5 (2014)

2014
[24]

The travelling salesman problem and adiabatic quantum computation: an algorithm,

T. D. Kieu, “The travelling salesman problem and adiabatic quantum computation: an algorithm,” Quantum Information Processing 18, 90 (2019)

2019
[25]

Randomized heuristics for the MAX-CUT problem,

P. Festa, P. M. Pardalos, M. G. Resende, and C. C. Ribeiro, “Randomized heuristics for the MAX-CUT problem,” Opti- mization methods and software 17, 1033–1058 (2002)

2002
[26]

Direct fidelity estimation from few Pauli measurements,

S. T. Flammia and Y .-K. Liu, “Direct fidelity estimation from few Pauli measurements,” Phys. Rev. Lett. 106, 230501 (2011)

2011
[27]

Sample- optimal tomography of quantum states,

J. Haah, A. W. Harrow, Z. Ji, X. Wu, and N. Yu, “Sample- optimal tomography of quantum states,” in Proceedings of the forty-eighth annual ACM symposium on Theory of Computing (2016) pp. 913–925

2016
[28]

Robust resource-efficient quantum variational ansatz through an evolutionary algorithm,

Y . Huang, Q. Li, X. Hou, R. Wu, M.-H. Yung, A. Bayat, and X. Wang, “Robust resource-efficient quantum variational ansatz through an evolutionary algorithm,” Phys. Rev. A 105, 052414 (2022)

2022
[29]

Quantum architec- ture search: a survey,

D. Martyniuk, J. Jung, and A. Paschke, “Quantum architec- ture search: a survey,” in 2024 IEEE International Confer- ence on Quantum Computing and Engineering (QCE) , V ol. 1 (IEEE, 2024) pp. 1695–1706

2024
[30]

MoG-VQE: Multiobjective genetic variational quantum eigensolver,

D. Chivilikhin, A. Samarin, V . Ulyantsev, I. Iorsh, A. R. Oganov, and O. Kyriienko, “MoG-VQE: Multiobjective genetic variational quantum eigensolver,” arxiv:2007.04424 (2020)

2020 arXiv
[31]

Adaptive diversity- based quantum circuit architecture search,

Y . Huang, S. Jin, B. Zeng, and Q. Shao, “Adaptive diversity- based quantum circuit architecture search,” Phys. Rev. Res.6, 033033 (2024)

2024
[32]

Barren plateaus in quantum neural network train- ing landscapes,

J. R. McClean, S. Boixo, V . N. Smelyanskiy, R. Babbush, and H. Neven, “Barren plateaus in quantum neural network train- ing landscapes,” Nat. Commun. 9, 4812 (2018)

2018
[33]

Cost function dependent barren plateaus in shal- low parametrized quantum circuits,

M. Cerezo, A. Sone, T. V olko ff, L. Cincio, and P. J. Coles, “Cost function dependent barren plateaus in shal- low parametrized quantum circuits,” Nat. Commun. 12, 1791 (2021)

2021
[34]

Barren plateaus in variational quantum comput- ing,

M. Larocca, S. Thanasilp, S. Wang, K. Sharma, J. Biamonte, P. J. Coles, L. Cincio, J. R. McClean, Z. Holmes, and M. Cerezo, “Barren plateaus in variational quantum comput- ing,” Nature Reviews Physics7, 174–189 (2025)

2025
[35]

An introduc- tion to quantum machine learning,

M. Schuld, I. Sinayskiy, and F. Petruccione, “An introduc- tion to quantum machine learning,” Contemporary Physics56, 172–185 (2015)

2015
[36]

Quantum machine learning,

J. Biamonte, P. Wittek, N. Pancotti, P. Rebentrost, N. Wiebe, and S. Lloyd, “Quantum machine learning,” Nature549, 195– 202 (2017)

2017
[37]

A comprehensive review of Quantum Machine Learning: from NISQ to Fault Tolerance,

Y . Wang and J. Liu, “A comprehensive review of Quantum Machine Learning: from NISQ to Fault Tolerance,” Reports on Progress in Physics (2024)

2024
[38]

Machine learning phases of matter,

J. Carrasquilla and R. G. Melko, “Machine learning phases of matter,” Nature Physics 13, 431–434 (2017)

2017
[39]

Learning phase transitions by confusion,

E. P. Van Nieuwenburg, Y .-H. Liu, and S. D. Huber, “Learning phase transitions by confusion,” Nature Physics 13, 435–439 (2017)

2017
[40]

Quantum gate learning in qubit networks: To ffoli gate without time-dependent con- trol,

L. Banchi, N. Pancotti, and S. Bose, “Quantum gate learning in qubit networks: To ffoli gate without time-dependent con- trol,” npj Quantum Information 2, 1–6 (2016)

2016
[41]

Machine-learning- assisted many-body entanglement measurement,

J. Gray, L. Banchi, A. Bayat, and S. Bose, “Machine-learning- assisted many-body entanglement measurement,” Phys. Rev. Lett. 121, 150503 (2018)

2018
[42]

Neural-network quantum state tomography,

G. Torlai, G. Mazzola, J. Carrasquilla, M. Troyer, R. Melko, and G. Carleo, “Neural-network quantum state tomography,” Nature physics 14, 447–450 (2018)

2018
[43]

Integrating neural networks with a quan- tum simulator for state reconstruction,

G. Torlai, B. Timar, E. P. Van Nieuwenburg, H. Levine, A. Omran, A. Keesling, H. Bernien, M. Greiner, V . Vuleti ´c, M. D. Lukin, et al., “Integrating neural networks with a quan- tum simulator for state reconstruction,” Phys. Rev. Lett. 123, 230504 (2019)

2019
[44]

Solving the quantum many-body problem with artificial neural networks,

G. Carleo and M. Troyer, “Solving the quantum many-body problem with artificial neural networks,” Science 355, 602– 606 (2017)

2017
[45]

Solving the Bose–Hubbard model with machine learning,

H. Saito, “Solving the Bose–Hubbard model with machine learning,” Journal of the Physical Society of Japan86, 093001 (2017)

2017
[46]

Multiclass classification of dephasing channels,

A. M. Palmieri, F. Bianchi, M. G. A. Paris, and C. Benedetti, “Multiclass classification of dephasing channels,” Phys. Rev. A 104, 052412 (2021)

2021
[47]

Quantum metrology assisted by machine learning,

J. Huang, M. Zhuang, J. Zhou, Y . Shen, and C. Lee, “Quantum metrology assisted by machine learning,” Advanced Quantum Technologies 8, 2300329 (2025)

2025
[48]

Bayesian Quan- tum Multiphase Estimation Algorithm,

V . Gebhart, A. Smerzi, and L. Pezz `e, “Bayesian Quan- tum Multiphase Estimation Algorithm,” Phys. Rev. Appl. 16, 014035 (2021)

2021
[49]

A machine learning ap- proach to Bayesian parameter estimation,

S. Nolan, A. Smerzi, and L. Pezz `e, “A machine learning ap- proach to Bayesian parameter estimation,” npj Quantum Infor- mation 7, 169 (2021). 11

2021
[50]

Unsupervised learning of non- Hermitian topological phases,

L.-W. Yu and D.-L. Deng, “Unsupervised learning of non- Hermitian topological phases,” Phys. Rev. Lett. 126, 240402 (2021)

2021
[51]

Nonuni- tary quantum machine learning,

J. Heredge, M. West, L. Hollenberg, and M. Sevior, “Nonuni- tary quantum machine learning,” Physical Review Applied23, 044046 (2025)

2025
[52]

Quantum computing models for artificial neural networks,

S. Mangini, F. Tacchino, D. Gerace, D. Bajoni, and C. Mac- chiavello, “Quantum computing models for artificial neural networks,” Europhysics Letters 134, 10002 (2021)

2021
[53]

Simulating a perceptron on a quantum computer,

M. Schuld, I. Sinayskiy, and F. Petruccione, “Simulating a perceptron on a quantum computer,” Physics Letters A 379, 660–663 (2015)

2015
[54]

Quantum circuit learning,

K. Mitarai, M. Negoro, M. Kitagawa, and K. Fujii, “Quantum circuit learning,” Physical Review A98 (2018)

2018
[55]

Data re-uploading for a universal quantum classi- fier,

A. P ´erez-Salinas, A. Cervera-Lierta, E. Gil-Fuster, and J. I. Latorre, “Data re-uploading for a universal quantum classi- fier,” Quantum 4, 226 (2020)

2020
[56]

E ffect of data encod- ing on the expressive power of variational quantum-machine- learning models,

M. Schuld, R. Sweke, and J. J. Meyer, “E ffect of data encod- ing on the expressive power of variational quantum-machine- learning models,” Physical Review A 103, 032430 (2021)

2021
[57]

Hierarchical quantum classifiers,

E. Grant, M. Benedetti, S. Cao, A. Hallam, J. Lockhart, V . Sto- jevic, A. G. Green, and S. Severini, “Hierarchical quantum classifiers,” npj Quantum Information 4 (2018)

2018
[58]

Quantum convolutional neural networks,

I. Cong, S. Choi, and M. D. Lukin, “Quantum convolutional neural networks,” Nature Physics 15, 1273–1278 (2019)

2019
[59]

Quan- volutional neural networks: powering image recognition with quantum circuits,

M. Henderson, S. Shakya, S. Pradhan, and T. Cook, “Quan- volutional neural networks: powering image recognition with quantum circuits,” Quantum Machine Intelligence2, 2 (2020)

2020
[60]

Absence of Barren Plateaus in Quantum Con- volutional Neural Networks,

A. Pesah, M. Cerezo, S. Wang, T. V olkoff, A. T. Sornborger, and P. J. Coles, “Absence of Barren Plateaus in Quantum Con- volutional Neural Networks,” Physical Review X11 (2021)

2021
[61]

Recurrent quantum neural networks,

J. Bausch, “Recurrent quantum neural networks,” Advances in neural information processing systems 33, 1368–1379 (2020)

2020
[62]

Harnessing disordered-ensemble quantum dynamics for machine learning,

K. Fujii and K. Nakajima, “Harnessing disordered-ensemble quantum dynamics for machine learning,” Physical Review Applied 8, 024030 (2017)

2017
[63]

Quantum reservoir processing,

S. Ghosh, A. Opala, M. Matuszewski, T. Paterek, and T. C. Liew, “Quantum reservoir processing,” npj Quantum Informa- tion 5, 35 (2019)

2019
[64]

Role of coherence in many-body Quantum Reservoir Computing,

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 Physics 7 (2024)

2024
[65]

Dissipation as a resource for Quantum Reser- voir Computing,

A. Sannia, R. Mart´ınez-Pe˜na, M. C. Soriano, G. L. Giorgi, and R. Zambrini, “Dissipation as a resource for Quantum Reser- voir Computing,” Quantum 8, 1291 (2024)

2024
[66]

Time-series quantum reservoir computing with weak and projective measurements,

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 Informa- tion 9 (2023)

2023
[67]

Dynamical Phase Transitions in Quantum Reservoir Computing,

R. Mart ´ınez-Pe˜na, G. L. Giorgi, J. Nokkala, M. C. Soriano, and R. Zambrini, “Dynamical Phase Transitions in Quantum Reservoir Computing,” Phys. Rev. Lett.127, 100502 (2021)

2021
[68]

Opportunities in quantum reservoir computing and extreme learning ma- chines,

P. Mujal, R. Mart´ınez-Pe˜na, J. Nokkala, J. Garc´ıa-Beni, G. L. Giorgi, M. C. Soriano, and R. Zambrini, “Opportunities in quantum reservoir computing and extreme learning ma- chines,” Advanced Quantum Technologies4, 2100027 (2021)

2021
[69]

Quantum ex- treme learning machines for photonic entanglement witness- ing,

D. Zia, L. Innocenti, G. Minati, S. Lorenzo, A. Suprano, R. D. Bartolo, N. Spagnolo, T. Giordani, V . Cimini, G. M. Palma, A. Ferraro, F. Sciarrino, and M. Paternostro, “Quantum ex- treme learning machines for photonic entanglement witness- ing,” arxiv:2502.18361 (2025)

2025
[70]

Experimen- tal Property Reconstruction in a Photonic Quantum Extreme Learning Machine,

A. Suprano, D. Zia, L. Innocenti, S. Lorenzo, V . Cimini, T. Giordani, I. Palmisano, E. Polino, N. Spagnolo, F. Sciarrino, G. M. Palma, A. Ferraro, and M. Paternostro, “Experimen- tal Property Reconstruction in a Photonic Quantum Extreme Learning Machine,” Phys. Rev. Lett.132,...

2024
[71]

Potential and limitations of quan- tum extreme learning machines,

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 Physics 6, 118 (2023)

2023
[72]

Quantum ensembles of quan- tum classifiers,

M. Schuld and F. Petruccione, “Quantum ensembles of quan- tum classifiers,” Scientific reports 8, 2772 (2018)

2018
[73]

Ensemble-learning error mitigation for variational quantum shallow-circuit classifiers,

Q. Li, Y . Huang, X. Hou, Y . Li, X. Wang, and A. Bayat, “Ensemble-learning error mitigation for variational quantum shallow-circuit classifiers,” Phys. Rev. Res.6, 013027 (2024)

2024
[74]

An e fficient quantum algorithm for ensemble classification using bagging,

A. Macaluso, L. Clissa, S. Lodi, and C. Sartori, “An e fficient quantum algorithm for ensemble classification using bagging,” IET Quantum Communication 5, 253–268 (2024)

2024
[75]

Training deep quantum neural networks,

K. Beer, D. Bondarenko, T. Farrelly, T. J. Osborne, R. Salz- mann, D. Scheiermann, and R. Wolf, “Training deep quantum neural networks,” Nat. Commun. 11 (2020)

2020
[76]

Train- ability of Dissipative Perceptron-Based Quantum Neural Net- works,

K. Sharma, M. Cerezo, L. Cincio, and P. J. Coles, “Train- ability of Dissipative Perceptron-Based Quantum Neural Net- works,” Phys. Rev. Lett.128 (2022)

2022
[77]

Power of data in quantum machine learning,

H.-Y . Huang, M. Broughton, M. Mohseni, R. Babbush, S. Boixo, H. Neven, and J. R. McClean, “Power of data in quantum machine learning,” Nat. Commun. 12, 2631 (2021)

2021
[78]

Practical quantum advantage in quantum simulation,

A. J. Daley, I. Bloch, C. Kokail, S. Flannigan, N. Pearson, M. Troyer, and P. Zoller, “Practical quantum advantage in quantum simulation,” Nature 607, 667–676 (2022)

2022
[79]

One qubit as a universal approxi- mant,

A. P ´erez-Salinas, D. L ´opez-N´u˜nez, A. Garc ´ıa-S´aez, P. Forn- D´ıaz, and J. I. Latorre, “One qubit as a universal approxi- mant,” Physical Review A104, 012405 (2021)

2021
[80]

Challenges and opportunities in quantum machine learning,

M. Cerezo, G. Verdon, H.-Y . Huang, L. Cincio, and P. J. Coles, “Challenges and opportunities in quantum machine learning,” Nature computational science 2, 567–576 (2022)

2022
[81]

Universal expressiveness of vari- ational quantum classifiers and quantum kernels for support vector machines,

J. J ¨ager and R. V . Krems, “Universal expressiveness of vari- ational quantum classifiers and quantum kernels for support vector machines,” Nat. Commun. 14, 576 (2023)

2023
[82]

Quantum Boltzmann Machine,

M. H. Amin, E. Andriyash, J. Rolfe, B. Kulchytskyy, and R. Melko, “Quantum Boltzmann Machine,” Physical Review X 8 (2018)

2018
[83]

Supervised quantum machine learning models are kernel methods,

M. Schuld, “Supervised quantum machine learning models are kernel methods,” arXiv:2101.11020 (2021)

2021 arXiv
[84]

Quantum phase recog- nition via quantum kernel methods,

Y . Wu, B. Wu, J. Wang, and X. Yuan, “Quantum phase recog- nition via quantum kernel methods,” Quantum 7, 981 (2023)

2023
[85]

Learning to classify quantum phases of matter with a few measurements,

M. Khosrojerdi, J. L. Pereira, A. Cuccoli, and L. Banchi, “Learning to classify quantum phases of matter with a few measurements,” Quantum Science and Technology10, 025006 (2025)

2025
[86]

Classically estimating ob- servables of noiseless quantum circuits,

A. Angrisani, A. Schmidhuber, M. S. Rudolph, M. Cerezo, Z. Holmes, and H.-Y . Huang, “Classically estimating ob- servables of noiseless quantum circuits,” arXiv2409.01706 (2024)

2024
[87]

Simulating quantum circuits with arbitrary local noise using pauli propagation,

A. Angrisani, A. A. Mele, M. S. Rudolph, M. Cerezo, and Z. Holmes, “Simulating quantum circuits with arbitrary local noise using pauli propagation,” arxiv:2501.13101 (2025)

2025 arXiv
[88]

Efficient Simulation of Parametrized Quan- tum Circuits under Nonunital Noise through Pauli Backprop- agation,

V . Martinez, A. Angrisani, E. Pankovets, O. Fawzi, and D. Stilck Franc ¸a, “Efficient Simulation of Parametrized Quan- tum Circuits under Nonunital Noise through Pauli Backprop- agation,” Phys. Rev. Lett.134, 250602 (2025)

2025
[89]

Quantum algo- rithms for supervised and unsupervised machine learning,

S. Lloyd, M. Mohseni, and P. Rebentrost, “Quantum algo- rithms for supervised and unsupervised machine learning,” arXiv:1307.0411 (2013)

2013 arXiv
[90]

Quantum advantage in learning from ex- periments,

H.-Y . Huang, M. Broughton, J. Cotler, S. Chen, J. Li, M. Mohseni, H. Neven, R. Babbush, R. Kueng, J. Preskill, and J. R. McClean, “Quantum advantage in learning from ex- periments,” Science 376, 1182–1186 (2022). 12

2022
[91]

Quantum Support Vector Machine for Big Data Classification,

P. Rebentrost, M. Mohseni, and S. Lloyd, “Quantum Support Vector Machine for Big Data Classification,” Phys. Rev. Lett. 113, 130503 (2014)

2014
[92]

Supervised learning with quantum-enhanced feature spaces,

V . Havl´ıˇcek, A. D. C ´orcoles, K. Temme, A. W. Harrow, A. Kandala, J. M. Chow, and J. M. Gambetta, “Supervised learning with quantum-enhanced feature spaces,” Nature 567, 209–212 (2019)

2019
[93]

Quantum state discrimination,

S. M. Barnett and S. Croke, “Quantum state discrimination,” Advances in Optics and Photonics 1, 238–278 (2009)

2009
[94]

Quantum Fingerprinting,

H. Buhrman, R. Cleve, J. Watrous, and R. de Wolf, “Quantum Fingerprinting,” Phys. Rev. Lett.87, 167902 (2001)

2001
[95]

Shadow tomography of quantum states,

S. Aaronson, “Shadow tomography of quantum states,” inPro- ceedings of the 50th annual ACM SIGACT symposium on the- ory of computing (2018) pp. 325–338

2018
[96]

Predicting many properties of a quantum system from very few measurements,

H.-Y . Huang, R. Kueng, and J. Preskill, “Predicting many properties of a quantum system from very few measurements,” Nature Physics 16, 1050–1057 (2020)

2020
[97]

Adam: A method for stochastic op- timization,

D. P. Kingma and J. Ba, “Adam: A method for stochastic op- timization,” arXiv:1412.6980 (2014)

2014 arXiv
[98]

Entanglement entropy scaling in solid-state spin arrays via capacitance measure- ments,

L. Banchi, A. Bayat, and S. Bose, “Entanglement entropy scaling in solid-state spin arrays via capacitance measure- ments,” Physical Review B94, 241117 (2016)

2016
[99]

Coherent Manipulation of Coupled Electron Spins in Semiconductor Quantum Dots,

J. R. Petta, A. C. Johnson, J. M. Taylor, E. A. Laird, A. Ya- coby, M. D. Lukin, C. M. Marcus, M. P. Hanson, and A. C. Gossard, “Coherent Manipulation of Coupled Electron Spins in Semiconductor Quantum Dots,” Science 309, 2180–2184 (2005)

2005
[100]

Radio frequency measurements of tunnel couplings and singlet–triplet spin states in si: P quantum dots,

M. House, T. Kobayashi, B. Weber, S. Hile, T. Watson, J. Van Der Heijden, S. Rogge, and M. Simmons, “Radio frequency measurements of tunnel couplings and singlet–triplet spin states in si: P quantum dots,” Nat. Commun. 6, 8848 (2015)

2015
[101]

Charge and spin state readout of a double quantum dot coupled to a resonator,

K. Petersson, C. Smith, D. Anderson, P. Atkinson, G. Jones, and D. Ritchie, “Charge and spin state readout of a double quantum dot coupled to a resonator,” Nano letters 10, 2789– 2793 (2010)

2010
[102]

Dipole Coupling of a Double Quantum Dot to a Microwave Resonator,

T. Frey, P. J. Leek, M. Beck, A. Blais, T. Ihn, K. Ensslin, and A. Wallraff, “Dipole Coupling of a Double Quantum Dot to a Microwave Resonator,” Phys. Rev. Lett.108, 046807 (2012)

2012
[103]

Dispersive Readout of a Few-Electron Double Quantum Dot with Fast rf Gate Sen- sors,

J. I. Colless, A. C. Mahoney, J. M. Hornibrook, A. C. Doherty, H. Lu, A. C. Gossard, and D. J. Reilly, “Dispersive Readout of a Few-Electron Double Quantum Dot with Fast rf Gate Sen- sors,” Phys. Rev. Lett.110, 046805 (2013)

2013
[104]

Preparation and Detection of Magnetic Quantum Phases in Optical Superlattices,

A. M. Rey, V . Gritsev, I. Bloch, E. Demler, and M. D. Lukin, “Preparation and Detection of Magnetic Quantum Phases in Optical Superlattices,” Phys. Rev. Lett.99, 140601 (2007)

2007
[105]

Time-Resolved Observation and Control of Su- perexchange Interactions with Ultracold Atoms in Optical Lat- tices,

S. Trotzky, P. Cheinet, S. F ¨olling, M. Feld, U. Schnorrberger, A. M. Rey, A. Polkovnikov, E. A. Demler, M. D. Lukin, and I. Bloch, “Time-Resolved Observation and Control of Su- perexchange Interactions with Ultracold Atoms in Optical Lat- tices,” Science 319, 295–299 (2008)

2008
[106]

Controlling and detecting spin correlations of ultra- cold atoms in optical lattices,

S. Trotzky, Y .-A. Chen, U. Schnorrberger, P. Cheinet, and I. Bloch, “Controlling and detecting spin correlations of ultra- cold atoms in optical lattices,” Phys. Rev. Lett. 105, 265303 (2010)

2010
[107]

The randomized measurement toolbox,

A. Elben, S. T. Flammia, H.-Y . Huang, R. Kueng, J. Preskill, B. Vermersch, and P. Zoller, “The randomized measurement toolbox,” Nature Reviews Physics5, 9–24 (2022)

2022
[108]

Determination of quasiprobability distributions in terms of probability distributions for the ro- tated quadrature phase,

K. V ogel and H. Risken, “Determination of quasiprobability distributions in terms of probability distributions for the ro- tated quadrature phase,” Phys. Rev. A40, 2847–2849 (1989)

1989
[109]

DeepSat: a learning framework for satellite imagery,

S. Basu, S. Ganguly, S. Mukhopadhyay, R. DiBiano, M. Karki, and R. Nemani, “DeepSat: a learning framework for satellite imagery,” Proceedings of the 23rd SIGSPATIAL International Conference on Advances in Geographic Infor- mation Systems , 1–10 (2015)

2015
[110]

MedMNIST Classification De- cathlon: A Lightweight AutoML Benchmark for Medical Im- age Analysis,

J. Yang, R. Shi, and B. Ni, “MedMNIST Classification De- cathlon: A Lightweight AutoML Benchmark for Medical Im- age Analysis,” in2021 IEEE 18th International Symposium on Biomedical Imaging (ISBI) (2021) pp. 191–195

2021
[111]

MedMNIST v2 - A large-scale lightweight bench- mark for 2D and 3D biomedical image classification,

J. Yang, R. Shi, D. Wei, Z. Liu, L. Zhao, B. Ke, H. Pfister, and B. Ni, “MedMNIST v2 - A large-scale lightweight bench- mark for 2D and 3D biomedical image classification,” Scien- tific Data 10, 41 (2023)

2023
[112]

Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algo- rithms,

H. Xiao, K. Rasul, and R. V ollgraf, “Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algo- rithms,” arXiv:1708.07747 (2017)

2017 arXiv
[113]

The MNIST Database of Handwritten Digit Images for Machine Learning Research [Best of the Web],

L. Deng, “The MNIST Database of Handwritten Digit Images for Machine Learning Research [Best of the Web],” IEEE Sig- nal Processing Magazine 29, 141–142 (2012). 13 SUPPLEMENTARY INFORMATION A. Data pre-processing For neural networks, the data pre-processing steps involve conve...

2012
[114]

In this study, we employ standard PyTorch nor- malization to shift and scale the image data to be in the range [−1, 1]

Data normalization: This preprocessing step ensures that the input features are on a similar scale, which can improve the convergence rate and generalization performance of the models. In this study, we employ standard PyTorch nor- malization to shift and scale the image data ...
[115]

Conversion to Tensors: The raw image data is converted into PyTorch tensors, which are multidimensional arrays that facilitate e fficient computations on GPUs or other hardware accelerators
[116]

Dataset Creation: The tensor data is organized into a Py- Torch Dataset, that provides a convenient way to access and iterate over the data samples during training and evaluation
[117]

DataLoader Creation: PyTorch DataLoaders are created from the Datasets, enabling e fficient batching, shu ffling, and parallel loading of data during the training process. B. Impact of the number of hidden layers To investigate the impact of network depth on classification per...

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.

[1] [1]

Algorithms for quantum computation: discrete logarithms and factoring,

P. W. Shor, “Algorithms for quantum computation: discrete logarithms and factoring,” in Proceedings 35th annual sym- posium on foundations of computer science (IEEE, 1994) pp. 124–134

1994

[2] [2]

A fast quantum mechanical algorithm for database search,

L. K. Grover, “A fast quantum mechanical algorithm for database search,” in Proceedings of the twenty-eighth annual ACM symposium on Theory of computing (1996) pp. 212–219

1996

[3] [3]

Quantum algo- rithm for linear systems of equations,

A. W. Harrow, A. Hassidim, and S. Lloyd, “Quantum algo- rithm for linear systems of equations,” Phys. Rev. Lett. 103, 150502 (2009)

2009

[4] [4]

Stabilizer Codes and Quantum Error Correc- tion,

D. Gottesman, “Stabilizer Codes and Quantum Error Correc- tion,” arXiv:9705052 (1997)

1997

[5] [5]

Theory of quantum error- correcting codes,

E. Knill and R. Laflamme, “Theory of quantum error- correcting codes,” Physical Review A55, 900 (1997)

1997

[6] [6]

Fault-tolerant quantum computation by anyons,

A. Y . Kitaev, “Fault-tolerant quantum computation by anyons,” Annals of physics 303, 2–30 (2003)

2003

[7] [7]

Noisy intermediate-scale quantum algo- rithms,

K. Bharti, A. Cervera-Lierta, T. H. Kyaw, T. Haug, S. Alperin- Lea, A. Anand, M. Degroote, H. Heimonen, J. S. Kottmann, T. Menke, et al. , “Noisy intermediate-scale quantum algo- rithms,” Reviews of Modern Physics94, 015004 (2022)

2022

[8] [8]

Quantum computing in the NISQ era and beyond,

J. Preskill, “Quantum computing in the NISQ era and beyond,” Quantum 2, 79 (2018)

2018