Pith. sign in

REVIEW 3 major objections 5 minor 24 references

Classical $\mathrm{SU}(2)$ Models Match or Exceed Shallow Variational Quantum Circuits on Vision Benchmarks

T0 review · 3 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read Classical quaternion networks match or beat shallow variational quantum circuits on vision benchmarks

desk verdict A careful, honest QML benchmark with a genuinely useful classical SU(2) baseline; the measurement bottleneck is a real caveat but the authors mostly own it. read the letter →

arxiv 2608.07822 v1 pith:2W26QCNP submitted 2026-08-07 cs.PF cs.LG

Christopher Fulton , Irene Tsapara , Lawrence Fulton This is my paper · ORCID
classification cs.PFcs.LG
keywords quaternionneuralnetworksvariationalquantumcircuitsSU(2)geometrymachinelearningvisionbenchmarksentanglementmeasurementbottlenecknaturalgradients
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

This paper asks whether the shared local SU(2) geometry of quaternion-valued neural networks and single-qubit variational quantum circuits translates into comparable learning behavior on classical image classification. On identical frozen features from MNIST, FashionMNIST, and CIFAR-10, quaternion classification heads match or closely approach real-valued MLPs while using fewer parameters, and they consistently outperform depth-3 product-state and entangled VQCs. The entangled circuit, which adds a ring of CNOT gates, gives only modest gains on grayscale data and degrades sharply under frozen ResNet18 features, falling 9.25 percentage points below the product-state circuit on CIFAR-10. The authors conclude that shared local SU(2) geometry and shallow ring entanglement are insufficient, within this shallow, measurement-limited regime, to confer practical quantum advantage on these tasks.

What carries the argument

The load-bearing object is the Lie group isomorphism between unit quaternions and SU(2): unit quaternions form the three-sphere $S^3$, which maps onto single-qubit rotations via a double cover of $SO(3)$. Quaternion layers implement this geometry as Hamilton products with renormalized unit-quaternion weights, while the VQCs implement the same local rotations as $R_Y$ and $R_Z$ gates with data re-uploading and an optional ring of CNOT gates for entanglement. The experimental design isolates this geometry by freezing feature extraction, a shared 16-dimensional learned bottleneck or a frozen 512-dimensional ResNet18 embedding, so that only the classification head differs; quantum heads then compress features into 4 or 8 qubits and read out 6 or 12 Pauli expectation values before a linear classifier.

What would settle it

Run the same frozen ResNet18 CIFAR-10 comparison with an amplitude-based encoding or a richer measurement strategy (e.g., classical shadow tomography or a much larger observable set) on the same depth-3 circuits; if the entangled VQC then matches or exceeds quaternion accuracy, the paper's conclusion that shallow SU(2) geometry and entanglement are insufficient would be refuted.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that a purely classical implementation of local SU(2) structure, quaternion-valued layers applying Hamilton products with unit-quaternion weights, reproduces the behavior of real-valued MLP heads and outperforms the shallow variational quantum classifiers evaluated here across all datasets and feature regimes. QuatNet matches RealNet within 0.2 percentage points on MNIST and FashionMNIST (93.64% vs 93.54% and 84.47% vs 84.60%) and retains 94–97% of real-valued accuracy on CIFAR-10, while the product-state VQC trails by 2.4–6.1 points and the entangled VQC drops 9.25 points below the product-state circuit under frozen ResNet18 features. The paper interprets this as evidence that measurement compression, shallow depth, and optimization instability, rather than the SU(2) rotation group itself, limit the quantum models.

Load-bearing premise

The central claim rests on the assumption that the particular shallow-VQC encoding and measurement scheme used here, compressing 512 features to 8 qubits and reading out only 12 Pauli observables, is representative enough of shallow variational quantum classifiers that its underperformance reflects a structural limit rather than an information bottleneck.

Editorial extensions

If this is right

  • Claims of practical quantum advantage on classical vision benchmarks should be benchmarked against parameter-efficient quaternion networks, not only real-valued MLPs, since both classical heads share the SU(2) local geometry.
  • The benefit of shallow entanglement depends on feature quality: it helps slightly on simple grayscale inputs, does nothing on a learned bottleneck, and actively hurts under rich pretrained CNN features.
  • Fubini–Study and quantum-Fisher-information natural gradients align with the natural-gradient direction but do not reduce short-horizon loss compared with Adam, indicating optimizer geometry is not the bottleneck.
  • Measurement compression is a prime suspect for the quantum gap: product-state circuits with 6 or 12 observables cannot carry the information that quaternion layers preserve in their continuous 4D representations.

Reading between the lines

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

  • A fair test of the paper's negative conclusion would require richer quantum encodings, such as amplitude encoding or larger observable sets (e.g., classical shadows), because the current design compresses 512 features into 8 qubits and 12 expectations, an information bottleneck the classical heads do not face.
  • The results suggest an empirical rule for near-term quantum machine learning: when classical data has no intrinsic quantum structure, the burden of proof shifts to showing that the encoding and measurement strategy, not the ansatz family, is what makes quantum features useful.
  • If the bottleneck explanation is right, one testable prediction is that enlarging the measurement space should narrow the gap between the entangled circuit and QuatNet on ResNet18 features, while deepening the circuit would mainly add barren-plateau difficulty.
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.

Referee Report

3 major / 5 minor

Summary. This paper presents a controlled benchmark comparing real-valued MLPs, quaternion-valued networks, and depth-3 variational quantum classifiers on identical frozen features from MNIST, FashionMNIST, and CIFAR-10, under both a 16-dimensional learned bottleneck and 512-dimensional frozen ResNet18 embeddings. Across all settings, QuatNet closely matches RealNet and outperforms both product-state and ring-entangled VQCs; the entangled circuit degrades sharply under pretrained CNN features. The authors add robustness checks (five-seed MNIST, sample scaling, an extra Quantum-NoEnt sample-size point) and FS/QFI optimization diagnostics. The paper concludes that shared local SU(2) geometry and shallow entanglement are insufficient for practical quantum advantage on these classical vision tasks, with conclusions explicitly bounded to the shallow, measurement-limited regime studied.

Significance. If the conclusion holds, the paper provides a useful classical baseline and a cautionary data point for quantum machine learning claims on classical vision benchmarks. Strengths include the frozen-feature protocol that isolates head architecture, explicit accuracy and variance tables, effect-size reporting, a five-seed MNIST extension, and an unusually frank limitations section. The main caveat is external validity: the VQC's 6- or 12-observable measurement bottleneck and the RealNet-trained learned bottleneck mean the headline claim is established for a narrow implementation, not for shallow VQCs generally. The paper's own text acknowledges the missing observable-count ablation, which is the key experiment needed to support the broader title claim.

major comments (3)
  1. [Quantum classifiers; Discussion, Scope, limitations] The 512D-to-8-qubit/12-observable compression is an information bottleneck that is not matched in the classical heads. In Table 2, Quantum-NoEnt uses 4,282 trainable parameters versus 17,832 for QuatNet and 66,954 for RealNet, and the final pre-classification representation has 12 real dimensions versus 128 hidden units for QuatNet. The authors state that expanding the measurement space would introduce an additional classical post-processing layer, and later acknowledge that the compression 'may disadvantage the quantum models.' This makes the title and abstract claim about 'shallow variational quantum circuits' broader than what the experiments establish. Please either add an experiment with a richer observable set (e.g., a full set of Pauli observables or classical-shadow-style measurements) or scope the title, abstract, and conclusion to 'shallow, measurement-limited variational quantum circuits.' This is load-bearing for the central claim.
  2. [Materials and Methods, Frozen feature extraction] The 16-dimensional learned bottleneck is 'trained once as part of the real-valued baseline (seed 42) and then frozen.' QuatNet and the VQC heads are therefore evaluated on features selected to optimize the real baseline's objective, which can systematically disadvantage the non-real heads in all MNIST, FashionMNIST, and 16-D CIFAR-10 experiments. The ResNet18 regime avoids this concern because those features are pretrained and not optimized for any head, but the bottleneck comparisons do not. Please either train the feature extractor jointly with each head, use a representation-neutral preprocessor for the bottleneck regime, or explicitly analyze how the real-baseline-trained extractor could bias the observed head ordering.
  3. [Discussion, Measurement bottlenecks and shallow-circuit limits] The paper concedes that 'the interpretation of performance gaps in terms of measurement compression and shallow-circuit expressivity limits ... has not been experimentally isolated through ablations on observable count or circuit expressivity.' This missing ablation is exactly what separates the effect of measurement compression from the effect of SU(2) geometry, entanglement, or circuit depth. Without it, the data support the narrow claim that the specific depth-3, 6/12-observable architectures tested underperform the classical heads; the broader explanatory claim that shared local SU(2) geometry and shallow entanglement are insufficient is not directly tested. Adding an observable-count sweep or a comparable-capacity classical head would strengthen the paper considerably.
minor comments (5)
  1. [Table 3] The single-seed entries at 2,000 and 2,500 samples per class for Quantum-Ent are described only in the table caption; please add an explicit footnote or marker in the table body itself so that readers do not mistake them for three-seed means.
  2. [Table 1 and Optimization geometry diagnostics] The FS/QFI diagnostic uses K=10 updates, and the text notes this is deliberately short-horizon. Please restate that horizon in the sentence summarizing the 'statistically indistinguishable' loss reduction, since the current wording in the Results could be read as a general optimizer equivalence claim.
  3. [Figure 5] The lower panels show log-scaled training time, but the axis label reads 'Training Time (seconds)' without indicating the log scale; please add '(log scale)' to the axis label.
  4. [Throughout] There are minor typographical issues, including 'F ashionMNIST' in Figure 5 labels and inconsistent spacing in 'T able 1' and 'T able 2'; these should be corrected in a final pass.
  5. [Robustness across seeds and sample sizes] The phrase that the Quantum-Ent degradation 'persists' under increased data exposure is based on single-seed runs at 2,000 and 2,500 samples, as the paper itself notes; consider using 'is observed at' or 'is indicative at' to reflect the reduced statistical strength of those points.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the paper is a self-contained empirical benchmark with independent model implementations, and its acknowledged information bottleneck is an external-validity limitation, not a circular derivation.

full rationale

This paper is an empirical benchmark, not a derivation from fitted constants, so the circularity burden is minimal. The central claim—that quaternion-valued heads match real-valued baselines and outperform depth-3 VQCs on identical frozen features—is established by training independently implemented models and evaluating them on held-out test sets. The SU(2) correspondence between quaternion layers and single-qubit gates is a mathematical isomorphism cited from standard external literature (Ref. [5]), not a relationship fitted to the experimental results. Both model families are implemented through separate code paths and parameterizations: QuatNet uses Hamilton products with renormalized unit-quaternion weights, while the VQCs use data re-uploading circuits with RY/RZ rotations and CNOT gates, so neither model's accuracy is an algebraic consequence of the other's definition. Frozen features are produced once by the real baseline but then used identically by all heads, meaning no head's performance is statistically forced by another head's fitted output. The trainable feature selector g_phi is a linear projection into rotation parameters, and no reported accuracy value is reconstructed from a fitted parameter renamed as a prediction. The paper explicitly acknowledges that the ResNet18 regime 'creates an information bottleneck that may affect the comparison between classical and quantum models' and that the interpretation of the gap 'has not been experimentally isolated through ablations on observable count or circuit expressivity.' These are honest external-validity caveats, not circular steps: they limit generalization of the conclusion without making the conclusion equivalent to its inputs. The reference list contains no load-bearing self-citations; all cited results (VQC architectures, barren plateaus, data re-uploading, quantum natural gradient) are external and independently verifiable. The conclusions are also explicitly bounded to 'shallow, measurement-limited variational circuits' on classical vision tasks, further reducing any overreach. Therefore no self-definitional, fitted-input-as-prediction, or self-citation circularity is present.

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

The paper introduces no analytical free parameters; the listed items are experimental design choices that the central claim depends on. The key assumptions are the fairness of the quantum encoding/measurement scheme and the frozen-feature comparability.

free parameters (5)
  • VQC circuit depth = 3 layers
    Chosen to preserve trainability; all VQC conclusions are bounded to this depth.
  • Number of measured observables = 6 (4-qubit), 12 (8-qubit)
    Measurement dimensionality fixed to isolate SU(2) geometry; this creates an information bottleneck for quantum heads.
  • Feature selector compression = 16D to 4 angles; 512D to 8 angles
    Trainable linear selector used to map frozen features to rotation angles; severe compression in the ResNet18 regime.
  • Adam learning rate = 1e-3
    Fixed across all models; no hyperparameter search reported for VQCs.
  • Learned bottleneck dimension = 16
    Low-capacity representation shared by all models in the bottleneck regime.
assumptions (5)
  • standard math Unit quaternions of norm 1 are isomorphic to SU(2) as Lie groups.
    Invoked in 'Unit quaternions and their correspondence to SU(2)' (Eq. for Phi) to motivate the comparison; standard result cited to Kuipers.
  • domain assumption Frozen feature representations with matched seeds isolate classification-head geometry.
    Materials and Methods, 'Frozen feature extraction' assumes performance differences derive from head inductive bias rather than representation quality.
  • domain assumption Noiseless GPU simulation adequately represents the variational circuits studied.
    All quantum results are simulated; the 'Scope, limitations' section notes hardware noise, decoherence, and finite-shot sampling are not captured.
  • ad hoc to paper The depth-3 ring-CNOT re-uploading circuit is representative of shallow VQCs.
    Conclusions generalize from one circuit family to 'shallow VQCs'; authors acknowledge richer entanglement schemes or encodings may alter results (Discussion).
  • domain assumption Small-sample statistics (n=5 Friedman, n=3 effect sizes) support the reported ordering.
    The paper uses asymptotic Friedman with n=5 and effect sizes as primary statistic for n=3; this inferential assumption is not independently justified.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Classical $\mathrm{SU}(2)$ Models Match or Exceed Shallow Variational Quantum Circuits on Vision Benchmarks." pith.science (2026). https://pith.science/paper/2W26QCNP

@misc{pith2026260807822,
  author       = {Pith},
  title        = {Pith review of: Classical $\mathrmSU(2)$ Models Match or Exceed Shallow Variational Quantum Circuits on Vision Benchmarks},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/2W26QCNP}},
  note         = {Machine review of arXiv:2608.07822}
}
abstract

Quaternion-valued neural networks and variational quantum circuits (VQCs) both derive local transformations from $\mathrm{SU}(2)$ geometry, yet their performance on classical supervised learning remains poorly understood. We compare real-valued, quaternion-valued, and quantum classification heads on identical frozen features across MNIST, FashionMNIST, and CIFAR-10. CIFAR-10 uses a learned 16-dimensional bottleneck and frozen ImageNet-pretrained ResNet18 features to separate architecture from representation quality. Quaternion classifiers match or approach real-valued baselines while outperforming shallow VQCs. On MNIST and FashionMNIST, quaternion networks nearly equal real-valued MLPs, whereas product-state VQCs show lower accuracy and higher cost. On CIFAR-10, quaternion networks retain 94--97% of real-valued performance and remain stable under a 32-fold increase in dimensionality. Product-state circuits underperform quaternion classifiers, while entanglement gives modest grayscale gains but reverses under pretrained CNN features (9.25 pp degradation vs.\ product-state). Fubini--Study/QFI natural gradients improve geometric alignment but not short-horizon loss reduction vs.\ Adam. A Friedman test on five-seed MNIST detects model differences ($\chi^2=12.796$, $p=0.0051$, $n=5$), with Wilcoxon tests yielding large effect sizes ($d>5$) for QuatNet vs.\ quantum comparisons. For FashionMNIST and CIFAR-10, large effects ($d>2.0$) are the primary statistic given $n=3$. These results indicate that quaternion networks provide efficient, stable $\mathrm{SU}(2)$ alternatives to shallow VQCs on tasks lacking intrinsic quantum structure. Shared local $\mathrm{SU}(2)$ geometry and shallow entanglement are insufficient, within the regime studied, to confer practical quantum advantage. Conclusions are limited to shallow, measurement-limited circuits on such tasks.

Figures

Figures reproduced from arXiv: 2608.07822 by the authors.

Figure 1
Figure 1. Schematic correspondence between unit quaternions and single-qubit SU(2) operations. Both parameterize transformations on S 3 , but quaternion layers preserve continuous four-dimensional representations through the Hamilton product, whereas VQCs apply unitary evolution followed by measurement. Contributions 1. Quaternion-valued SU(2) models closely match real-valued baselines while outperforming shallow VQCs. Across… view at source ↗
Figure 2
Figure 2. Experimental pipeline overview. Models are compared under matched data, feature, and training conditions across three datasets and two feature regimes: a 16-dimensional learned bottleneck and frozen 512-dimensional ResNet18 embeddings. Datasets and preprocessing We evaluate three standard image-classification benchmarks: MNIST handwritten digits [12], FashionMNIST clothing items [13], and CIFAR-10 natural images [14… view at source ↗
Figure 3
Figure 3. Depth-3 VQC architectures used in the four-qubit MNIST/FashionMNIST regime. Panel (a) shows the product-state model (Quantum-NoEnt), and panel (b) adds a ring of CNOT gates (Quantum-Ent). Each circuit uses data re-uploading with trainable rotations and outputs six observables passed to a linear classifier. isolate gradient geometry from downstream representation effects. The goal was not to optimize accuracy, but to… view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Architecture comparison of real-valued, quaternion, and quantum classifiers under shared frozen features. Quaternion heads apply Hamilton products with unit quaternions, while quantum heads use depth-3 variational circuits with and without ring-topology CNOT entangleme…
Figure 5
Figure 5. Figure 5: Performance across three random seeds on MNIST, FashionMNIST, and CIFAR-10 under two frozen feature regimes. Top panels show test accuracy, and bottom panels show log-scaled training time. Bars show seed ranges and markers denote means. Robustness across seeds and samp…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

24 extracted references · 15 canonical work pages

  1. [1]

    Artificial Intelligence Review53, 2957–2982 (2020) https://doi.org/10.1007/ s10462-019-09752-1

    Parcollet, T., Morchid, M., Linarès, G.: A survey of quaternion neural networks. Artificial Intelligence Review53, 2957–2982 (2020) https://doi.org/10.1007/ s10462-019-09752-1

    work page 2020

  2. [2]

    In: Proceedings of the International Conference on Learning Representations (ICLR) (2019)

    Parcollet, T., Ravanelli, M., Morchid, M., Linarès, G., Trabelsi, C., De Mori, R., Bengio, Y.: Quaternion Recurrent Neural Networks. In: Proceedings of the International Conference on Learning Representations (ICLR) (2019)

    work page 2019

  3. [3]

    In: Proceedings of the IEEE International Joint Conference on Neural Networks (IJCNN), pp

    Gaudet, C.J., Maida, A.S.: Deep Quaternion Networks. In: Proceedings of the IEEE International Joint Conference on Neural Networks (IJCNN), pp. 1–8 (2018). IEEE

    work page 2018

  4. [4]

    John Wiley & Sons, Chichester, UK (2009)

    Mandic, D., Goh, V.S.L.: Complex Valued Nonlinear Adaptive Filters: Noncir- cularity, Widely Linear and Neural Models. John Wiley & Sons, Chichester, UK (2009)

    work page 2009

  5. [5]

    Princeton University Press, Princeton, NJ (1999)

    Kuipers, J.B.: Quaternions and Rotation Sequences: A Primer with Applications to Orbits, Aerospace, and Virtual Reality. Princeton University Press, Princeton, NJ (1999)

    work page 1999

  6. [6]

    Physical Review A101(3), 032308 (2020)

    Schuld, M., Bocharov, A., Svore, K.M., Wiebe, N.: Circuit-Centric Quantum Classifiers. Physical Review A101(3), 032308 (2020)

    work page 2020

  7. [7]

    Nature Communications 11(1), 808 (2020)

    Beer,K.,Bondarenko,D.,Farrelly,T.,Osborne,T.J.,Salzmann,R.,Scheiermann, D., Wolf, R.: Training Deep Quantum Neural Networks. Nature Communications 11(1), 808 (2020)

    work page 2020

  8. [8]

    Advanced Quantum Technologies2(12), 1900070 (2019) https://doi.org/10.1002/ qute.201900070

    Sim, S., Johnson, P.D., Aspuru-Guzik, A.: Expressibility and entangling capabil- ity of parameterized quantum circuits for hybrid quantum-classical algorithms. Advanced Quantum Technologies2(12), 1900070 (2019) https://doi.org/10.1002/ qute.201900070

    work page 2019

Show all 24 references
  1. [9]

    Nature567(7747), 209–212 (2019) https://doi.org/10.1038/ s41586-019-0980-2 24

    Havlíček, V., Córcoles, A.D., Temme, K., Harrow, A.W., Kandala, A., Chow, J.M., Gambetta, J.M.: Supervised learning with quantum-enhanced feature spaces. Nature567(7747), 209–212 (2019) https://doi.org/10.1038/ s41586-019-0980-2 24

    2019

  2. [10]

    arXiv preprint arXiv:1811.04968 (2018)

    Bergholm, V., Izaac, J., Schuld, M., Gogolin, C., Ahmed, S., Ajith, V., Alam, M.S., Alonso-Linaje, G., Narayanan, B.A., Asadi, A., et al.: PennyLane: Auto- matic differentiation of hybrid quantum–classical computations. arXiv preprint arXiv:1811.04968 (2018)

    2018 arXiv

  3. [11]

    https://developer.nvidia.com/cuquantum-sdk (2021)

    NVIDIA Corporation: cuQuantum SDK: A High-Performance Library for Quantum Circuit Simulation. https://developer.nvidia.com/cuquantum-sdk (2021)

    2021

  4. [12]

    Proceedings of the IEEE86(11), 2278–2324 (1998)

    LeCun, Y., Bottou, L., Bengio, Y., Haffner, P.: Gradient-based learning applied to document recognition. Proceedings of the IEEE86(11), 2278–2324 (1998)

    1998

  5. [13]

    arXiv preprint arXiv:1708.07747 (2017)

    Xiao, H., Rasul, K., Vollgraf, R.: Fashion-mnist: a novel image dataset for benchmarking machine learning algorithms. arXiv preprint arXiv:1708.07747 (2017)

    2017 arXiv

  6. [14]

    Technical Report CSRG-TR-2009-1, University of Toronto, Toronto, Canada (2009)

    Krizhevsky, A.: Learning multiple layers of features from tiny images. Technical Report CSRG-TR-2009-1, University of Toronto, Toronto, Canada (2009)

    2009

  7. [15]

    Nature Communica- tions9(1), 4812 (2018) https://doi.org/10.1038/s41467-018-07090-4

    McClean, J.R., Boixo, S., Smelyanskiy, V.N., Babbush, R., Neven, H.: Barren plateaus in quantum neural network training landscapes. Nature Communica- tions9(1), 4812 (2018) https://doi.org/10.1038/s41467-018-07090-4

    2018 doi

  8. [16]

    Quantum Science and Technology6(2), 025011 (2021) https: //doi.org/10.1088/2058-9565/abbf6b arXiv:2008.02785

    Huembeli, P., Dauphin, A.: Characterizing the loss landscape of variational quantum circuits. Quantum Science and Technology6(2), 025011 (2021) https: //doi.org/10.1088/2058-9565/abbf6b arXiv:2008.02785

    2021 arXiv

  9. [17]

    Schuld, M., Killoran, N.: Is quantum advantage the right goal for quantum machine learning? PRX Quantum3(3), 030101 (2022) https://doi.org/10.1103/ PRXQuantum.3.030101

    2022

  10. [18]

    Quantum4, 226 (2020) https: //doi.org/10.22331/q-2020-02-06-226

    Pérez-Salinas, A., Cervera-Lierta, A., Gil-Fuster, E., Latorre, J.I.: Data re- uploading for a universal quantum classifier. Quantum4, 226 (2020) https: //doi.org/10.22331/q-2020-02-06-226

    2020 doi

  11. [19]

    Quantum4, 269 (2020) https://doi.org/10.22331/q-2020-05-25-269

    Stokes, J., Izaac, J., Killoran, N., Carleo, G.: Quantum natural gradient. Quantum4, 269 (2020) https://doi.org/10.22331/q-2020-05-25-269

    2020 doi

  12. [20]

    Princeton University Press, Princeton, NJ (2008)

    Absil, P.-A., Mahony, R., Sepulchre, R.: Optimization Algorithms on Matrix Manifolds. Princeton University Press, Princeton, NJ (2008)

    2008

  13. [21]

    PRX Quantum3(1), 010313 (2022) https://doi.org/10.1103/PRXQuantum.3.010313

    Holmes, Z., Sharma, K., Cerezo, M., Coles, P.J.: Connecting ansatz expressibility to gradient magnitudes and barren plateaus. PRX Quantum3(1), 010313 (2022) https://doi.org/10.1103/PRXQuantum.3.010313

    2022 doi

  14. [22]

    Nature Communications12(1), 2631 (2021) https://doi.org/10.1038/ 25 s41467-021-22539-9

    Huang, H.-Y., Kueng, R., Preskill, J.: Power of data in quantum machine learning. Nature Communications12(1), 2631 (2021) https://doi.org/10.1038/ 25 s41467-021-22539-9

    2021

  15. [23]

    arXiv preprint arXiv:2311.05559 (2023) arXiv:2311.05559 [quant-ph]

    Kölle, M., Maurer, J., Altmann, P., Sünkel, L., Stein, J., Linnhoff-Popien, C.: Disentangling quantum and classical contributions in hybrid quantum machine learning architectures. arXiv preprint arXiv:2311.05559 (2023) arXiv:2311.05559 [quant-ph]

    2023 arXiv

  16. [24]

    Quantum2, 79 (2018) https://doi.org/10.22331/q-2018-08-06-79 26

    Preskill, J.: Quantum computing in the NISQ era and beyond. Quantum2, 79 (2018) https://doi.org/10.22331/q-2018-08-06-79 26

    2018 doi

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.