Pith. sign in

REVIEW 3 major objections 6 minor 27 references

Quantum Circuit Training with Growth-Based Architectures

T0 review · 3 major / 6 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read The paper claims that growing a parameterized quantum circuit during training—adding identity-initialized reuploader or feature-map blocks when needed—yields lower losses, less variance, and better generalization than fixed-depth training.

desk verdict Useful empirical study of growing PQCs, but the 'final loss' claim is undercut by best-epoch test-set selection and undisclosed growth schedules. read the letter →

arxiv 2411.16560 v1 pith:37UBYFPX submitted 2024-11-25 quant-ph cs.LG

Callum Duffy , Smit Chaudhary , Gergana V. Velikova This is my paper · ORCID
classification quant-phcs.LG
keywords parameterizedquantumcircuitsgrowth-basedarchitecturesreuploaderfrequencyspectrummachinelearningbarrenplateausoverfitting2DLaplaceequation
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

Training a parameterized quantum circuit (PQC) usually means fixing its architecture and optimizing all parameters from the start. This paper argues that starting small and letting the circuit grow during training is better. Three growth strategies—Block Growth, Sequential Feature Map Growth, and Interleave Feature Map Growth—insert reuploader blocks or feature-map gates one stage at a time, initialized as identities so the function already learned is preserved. Because a reuploader circuit—one that alternates data-encoding feature maps with trainable ansatz blocks—is a truncated Fourier series, each added feature map widens the accessible frequency spectrum, so capacity tracks what the data actually needs. Across student-teacher regressions, a noisy regression, and the 2D Laplace equation, the growing circuits achieve lower final losses, lower variance across random seeds, and better generalization than fixed-depth complete-depth learning, including a regime where the fixed circuit appears to hit a barren plateau.

What carries the argument

The load-bearing object is the reuploader PQC viewed as a truncated Fourier series, $f(x,\theta,\psi) = \sum_{\omega_j \in \Omega} c_j(\theta,\hat{C}) e^{i \omega_j(\psi)\cdot \phi(x)}$. The accessible frequencies $\Omega$ come from the eigenspectrum of the feature-map generator $\hat{G} = \sum_m \psi_m \hat{X}_m/2$; repeating the feature map more times enlarges $|\Omega|$ roughly linearly. Growth methods insert identity-initialized blocks: each added gate is paired with a cancellation gate so the circuit still implements the identity at insertion, preserving the learned function while giving the new parameters nonzero gradients from the start. Block Growth appends whole reuploader blocks (feature map plus ansatz), increasing both the frequency range and the Fourier-coefficient expressivity; Sequential FM Growth inserts feature maps from left to right between existing ansatz blocks; Interleave FM Growth inserts them from the center outward. Identity insertion is what makes mid-training growth safe, and the Fourier-spectrum view turns growth into a controlled expansion of capacity rather than a random widening.

What would settle it

Rerun the student-teacher and Laplace experiments with a sweep over growth schedules (epochs between growth, number of blocks $\ell$ per stage, stopping patience), selecting schedules by validation loss for each method; if fixed-depth CDL with the same final depth and identical initialization then matches or beats the best growth model, the reported advantage is a scheduling artifact rather than a property of growth.

Watch

Extended reading notes

Core claim

The paper's central discovery is that dynamically grown PQCs dominate static ones in every setting tested. In student-teacher tasks where a randomly initialized teacher circuit with $L$ reuploader layers generates the data, all three growth methods reach the same final depth as the teacher but end with lower mean and worst-case mean squared error than complete-depth learning: the best 1-qubit mean is $(1.029 \pm 3.52)\times 10^{-6}$ for Interleave FM Growth versus $(8.13 \pm 1.42)\times 10^{-5}$ for the best randomly initialized CDL baseline. On a noisy 20-point regression, block growth finishes with train/test losses of $0.035/0.063$, while a 20-layer CDL model overfits to $0.005/0.158$. Solving the 2D Laplace equation with 5 to 13 reuploader layers, all three growth methods keep low $L^2$ relative error with small variance, while identity-initialized CDL degrades sharply at 13 layers—behavior the paper attributes to the onset of a barren plateau. The common mechanism is spectral growth: expanding the model's Fourier frequencies gradually acts as a regularizer and keeps the circuit shallow until extra capacity is justified.

Load-bearing premise

The central comparison assumes that the growth schedule—how many epochs pass before growing, how many blocks are added per stage, and when growth stops—was fixed in advance and not tuned on the test set to favor the growing methods.

Editorial extensions

If this is right

  • Practitioners can start with a compact PQC and grow it only as the task demands, keeping circuits shallower on noisy hardware without sacrificing final accuracy.
  • Growth acts as an implicit regularizer: because high-frequency modes are added late, models trained on small noisy datasets are less likely to fit noise, as shown by the noisy student-teacher train/test gap.
  • At depths where fixed-depth training degrades (the 13-layer Laplace case), growth-based training maintains performance, which the authors interpret as resilience to barren-plateau onset.
  • The three growth variants are largely interchangeable in accuracy on the Laplace task, so the choice can be driven by hardware layout or implementation convenience rather than performance.
  • Identity initialization of added blocks preserves the current function, so growth can be inserted into an ongoing optimization without resetting progress.

Reading between the lines

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

  • Extension: if the spectral explanation is right, growth should act as a curriculum over Fourier modes—low frequencies are learned first and higher frequencies added only when the target demands them—so the gains should be largest on targets with decaying spectral power, such as natural images or PDE solutions.
  • Extension: the paper triggers growth after a fixed number of epochs; an adaptive trigger based on validation-loss plateaus would make the method self-terminating and would likely enlarge its advantage on tasks of unknown complexity.
  • Extension: a decisive comparison not reported here is against a fixed-depth CDL model matched in total final parameters; if growth still wins at equal final capacity, the benefit comes from the training trajectory itself rather than from having fewer parameters.
  • Extension: because added feature maps enlarge the frequency spectrum roughly linearly in repetitions, the approach could be combined with trainable frequency generators (the paper's Equation 3 setting) to give the optimizer continuous control over both coefficients and spectrum.
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 / 6 minor

Summary. This paper proposes three growth-based training strategies for reuploader parameterized quantum circuits (Block Growth, Sequential Feature Map Growth, and Interleave Feature Map Growth), in which identity-initialized reuploader blocks are added during training to expand the accessible frequency spectrum. The methods are compared with fixed-depth complete-depth learning (CDL) circuits, with both random and identity initialization, on three tasks: student-teacher regression for 1- and 2-qubit circuits, a noisy student-teacher regression, and a 2D Laplace equation benchmark with an analytic solution. The paper reports that growth-based methods achieve lower losses, reduced variance across 50 random seeds, and better generalization than fixed-depth approaches, and it provides a pseudocode training procedure and a link to a code repository.

Significance. The central idea of growing a PQC during training to balance expressivity and trainability is timely and practically relevant for quantum scientific machine learning. The paper has several strengths: the use of 50 random seeds, clear boxplot visualizations, an analytic Laplace solution for a physically meaningful benchmark, a noisy regression setting that directly tests overfitting, and a publicly available code repository. If the reported advantages hold under a correct evaluation protocol, the work would be a useful contribution to adaptive variational quantum circuit training. However, the central claim is currently overstated because the main student-teacher experiments report best-epoch test losses while the abstract claims lower final losses, and the growth schedule hyperparameters are not disclosed. These issues undermine the current evidence for the paper's headline conclusions, although they appear fixable with additional reporting and analysis.

major comments (3)
  1. [Section 3.1 and Algorithm 1; Abstract] The abstract claims that growth-based PQCs achieve 'lower final losses and reduced variance between runs,' but the primary student-teacher experiments in Section 3.1 report 'the best loss achieved during each run,' and Algorithm 1 evaluates the test loss every epoch without storing or reporting the final-epoch loss. Tables 1-2 and Figures 5a-5b therefore summarize distributions of best test losses selected over the training trajectory, not final losses. Selecting the best epoch on the test set is a form of test-set model selection and can change the relative ranking of methods with different loss trajectories; for example, a growing method that peaks early and then degrades would look better than a fixed-depth method that converges slowly but ends lower. Because the abstract's central claim is about final losses and Section 3.1 is the main controlled comparison, this mismatch is load-bearing. The authors should either report final-epoch losses (or use a validation-based early-stopping rule) or explicitly revise the claim to refer to best-attained losses rather than final losses.
  2. [Section 2 and Section 3.2] The growth hyperparameters are never specified. Section 2 says circuits grow 'after a predefined number of epochs' and that the number of blocks added per stage, ℓ, is 'a user-defined number,' while Section 3.2 says growth continues 'until no further improvement is observed.' No values for the growth interval, ℓ, patience, or stopping threshold are given anywhere in the paper. Without these details the experiments are not reproducible, and the reader cannot rule out that the reported advantage of growth over CDL depends on task-specific tuning of the growth schedule, including possible selection based on test performance. The authors should disclose the exact schedule, all growth-related hyperparameters, and the criterion used to decide when to stop growing, ideally with a sensitivity analysis.
  3. [Appendix A.2 and Section 3.1.2] In the 2-qubit student-teacher task, the reported mean best losses for the growth methods are 3.19e-4, 3.21e-4, and 3.75e-4, while the 9-layer identity-initialized CDL model has mean 3.97e-4 with standard errors of roughly 1.4e-4 to 3.0e-4 for all methods. The differences between the growth methods and the best identity-initialized CDL baseline are therefore small relative to the reported standard errors, and no statistical significance tests are provided. The text states that the growing models 'outperformed both the identity-initialized CDL circuits,' but the reported summary statistics do not clearly establish this. The authors should add appropriate hypothesis tests or confidence intervals, or temper the claim in light of the overlapping uncertainties.
minor comments (6)
  1. [Section 2, Eq. (1)] There is a typo in 'generator Hamiltonian’s' and the notation Ġ_m is used inconsistently: Eq. (1) writes Ġ_m(γ_m), while Eq. (3) and the following paragraph use Ġ_m(γ_m, ψ).
  2. [Figures 1-3] The circuit diagrams are not self-contained; the captions should define whether U_i and F_i represent ansatz blocks, feature map blocks, or both, and how they correspond to the growth stages described in the text.
  3. [Figure 5] The labels '5-layer CDL ()' and '20-layer CDL ()' appear to have a missing symbol, likely the identity-initialization marker; please fix the rendering.
  4. [Section 3.3, Figure 8] The x-axis label and caption for Figure 8 are ambiguous for growth methods: 'began with or grew to' does not clarify whether the reported layer count is the initial depth, the final depth, or a target depth for each method.
  5. [References] Reference [6] is incomplete (no title or journal details), Reference [8] is a bare Semantic Scholar URL rather than a full citation, and References [9] and [21] cite the same paper by Grant et al. with different levels of detail; these should be unified.
  6. [Algorithm 1] Algorithm 1 computes the test loss each epoch but does not state how the 'best loss' reported in Section 3.1 is obtained; please make explicit the selection procedure and clarify whether it uses the test set or a validation set.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found: the paper is an empirical method comparison with measured losses and errors, not a derivation that reduces to its own inputs.

full rationale

The paper's central claims are empirical: growing PQCs are compared with fixed-depth CDL baselines on student-teacher regression, a noisy regression task, and a 2D Laplace PDE, using measured MSE and L2 relative error over 50 seeds. The Fourier-series description in Equations 1-3 is cited from prior external work (Schuld et al., Ref. [17]) and is used only as motivation for why adding feature-map blocks expands the accessible frequency spectrum; it is not the source of the reported numerical results. No parameter is fitted to a subset of data and then renamed as a prediction: the identity-initialized growth procedure is a training algorithm, and all reported values are direct training/test losses. There are no load-bearing self-citations: the references to prior initialization and layerwise-learning work (Refs. [18], [21]) are independent published results, not claims by the present authors, and the paper does not invoke any uniqueness theorem or ansatz smuggled in via citation. The only notable internal issue is that Section 3.1 reports the best loss achieved during each run while the Abstract claims lower final losses, and selecting the best test epoch is a metric/reporting concern rather than a circular-reasoning defect. Under the required definitions, the derivation chain is therefore self-contained and not circular.

Assumptions & free parameters 4 free parameters · 6 assumptions · 0 invented entities

The central empirical claims rest on the standard Fourier characterization of reuploader circuits plus several domain assumptions about low-frequency dominance, identity initialization, and convergence of the training runs. The main untracked degrees of freedom are the growth schedule, ℓ, and stopping criterion, none of which are reported. No new physical or mathematical entities are introduced.

free parameters (4)
  • growth_epoch_interval = not stated
    Section 2 says growth happens 'after a predefined number of epochs' but no value is given; this timing is central to all experiments.
  • blocks_per_growth_stage_l = not stated
    Section 2 calls ℓ a 'user-defined number' of blocks added per growth stage; values are not reported in Section 3, and this controls the final depth and expressivity growth rate.
  • growth_stopping_criterion = not stated
    Section 3.2 says growth continues 'until no further improvement is observed' but no patience, threshold, or validation metric is specified.
  • teacher_initialization_ranges = 1-qubit ansatz [0,0.1], FM [0,pi/9]; 2-qubit [0,pi/5]
    The teacher circuit outputs in Section 3.1 are generated from these ranges; task difficulty and relative method rankings could depend on them.
assumptions (6)
  • standard math A reuploader PQC output equals a truncated Fourier series whose frequencies are determined by the eigenspectrum of the generator G.
    Used in Section 2, Eqs. (1)-(3), relying on Schuld et al. [17]; this is the foundation for the frequency-growth interpretation.
  • standard math Adding reuploader blocks or feature-map repetitions increases the size of the accessible frequency spectrum.
    Stated in Section 2 as a linear dependence of spectrum size on the number of encoding repetitions; underpins why growth changes model capacity.
  • domain assumption Natural and PDE data are dominated by low-frequency components, so limiting high-frequency access acts as a regularizer.
    Invoked in Section 3.2, citing [22,23]; this is the mechanism claimed for the overfitting-mitigation result.
  • domain assumption Identity-initialized paired gates preserve the currently represented function and provide nonzero gradients from the start of training.
    Stated in Section 2, based on Grant et al. [21]; required so that growth does not destroy learned behavior.
  • domain assumption In the student-teacher setup, when the student reaches the teacher's depth, the teacher's function lies exactly in the student's model class.
    Section 3.1 uses this to interpret losses as a pure measure of optimization quality.
  • domain assumption Adam with the stated learning rates and 1000 or 2000 epochs gives converged training for all compared models.
    Section 3.3 asserts convergence because losses plateaued within the epoch budget; no formal convergence check is provided.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Quantum Circuit Training with Growth-Based Architectures." pith.science (2026). https://pith.science/paper/37UBYFPX

@misc{pith2026241116560,
  author       = {Pith},
  title        = {Pith review of: Quantum Circuit Training with Growth-Based Architectures},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/37UBYFPX}},
  note         = {Machine review of arXiv:2411.16560}
}
read the original abstract

This study introduces growth-based training strategies that incrementally increase parameterized quantum circuit (PQC) depth during training, mitigating overfitting and managing model complexity dynamically. We develop three distinct methods: Block Growth, Sequential Feature Map Growth, and Interleave Feature Map Growth, which add reuploader blocks to PQCs adaptively, expanding the accessible frequency spectrum of the model in response to training needs. This approach enables PQCs to achieve more stable convergence and generalization, even in noisy settings. We evaluate our methods on regression tasks and the 2D Laplace equation, demonstrating that dynamic growth methods outperform traditional, fixed-depth approaches, achieving lower final losses and reduced variance between runs. These findings underscore the potential of growth-based PQCs for quantum scientific machine learning (QSciML) applications, where balancing expressivity and stability is essential.

Figures

Figures reproduced from arXiv: 2411.16560 by the authors.

Figure 1
Figure 1. PQC growing method block growth where both feature maps and ansatzs blocks are added dur￾ing training, with some specified block being added ℓ times every time the PQC grows. U0 F0 U1 F1 U2 F2 U3 [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Sequential feature map growth, where fea [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Interleave feature map growth, where fea [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: Teacher circuit architecture for 2 qubits, randomly initialized to create a dataset for PQCs learn [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: Mean squared error performance for different training methods on the student-teacher regression [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]
Figure 6
Figure 6. Figure 6: The mean square error of both the 20- layer CDL and block growth models on the 1D noisy student-teacher model on a log-scale over the course of training for 1000 epochs. Reporting the mean train and test error as the solid lines and the standard de￾viation on the mean …
Figure 7
Figure 7. Figure 7: The outputs of PQCs trained on the noisy 1D student-teacher dataset. Depicting both 20-layer [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]
Figure 8
Figure 8. Figure 8: L2 relative error of various models solv￾ing the 2D Laplace equation. The plot compares the mean L2 relative error in bold with standard error shaded for each model and how many reuploader lay￾ers it began with or grew to over 50 training runs. Models include complete-…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

27 extracted references · 14 canonical work pages

  1. [18]

    Layerwise learning for quantum neural networks,

    A. Skolik, J. R. McClean, M. Mohseni, P. van der Smagt, and M. Leib, “Layerwise learning for quantum neural networks,” Quantum Ma- chine Intelligence, vol. 3, no. 1, Jan. 2021, issn: 2524-4914. doi: 10.1007/s42484-020-00036-

    work page doi:10.1007/s42484-020-00036- 2021

  2. [1]

    Farhi, J

    E. Farhi, J. Goldstone, and S. Gutmann, A quantum approximate optimization algorithm,

    work page

  3. [2]

    Verdon et al., Learning to learn with quan- tum neural networks via classical neural net- works, 2019

    G. Verdon et al., Learning to learn with quan- tum neural networks via classical neural net- works, 2019. arXiv: 1907.05415 [quant-ph]. [Online]. Available: https://arxiv.org/abs/ 1907.05415

    arXiv 2019

  4. [3]

    From the quantum approximate optimization algorithm to a quantum alternating operator ansatz,

    S. Hadfield, Z. Wang, B. O’Gorman, E. G. Ri- effel, D. Venturelli, and R. Biswas, “From the quantum approximate optimization algorithm to a quantum alternating operator ansatz,” Al- gorithms, vol. 12, no. 2, p. 34, Feb. 2019, issn: 1999-4893. doi: 10.3390/a12020034. [Online]. Available: http : / / dx . doi . org / 10 . 3390 / a12020034

    doi:10.3390/a12020034 2019

  5. [4]

    Quantum approximate optimization al- gorithm for maxcut: A fermionic view,

    Z. Wang, S. Hadfield, Z. Jiang, and E. G. Ri- effel, “Quantum approximate optimization al- gorithm for maxcut: A fermionic view,” Phys. Rev. A, vol. 97, p. 022 304, 2 2018. doi: 10 . 1103 / PhysRevA . 97 . 022304. [Online]. Avail- able: https://link.aps.org/doi/10.1103/ PhysRevA.97.022304

    work page 2018

  6. [5]

    Training the quantum approximate optimization algorithm without access to a quantum processing unit,

    M. Streif and M. Leib, “Training the quantum approximate optimization algorithm without access to a quantum processing unit,” Quan- tum Science and Technology, vol. 5, no. 3, p. 034 008, 2020. doi: 10 . 1088 / 2058 - 9565 / ab8c2b. [Online]. Available: https://dx.doi. org/10.1088/2058-9565/ab8c2b

    doi:10.1088/2058-9565/ab8c2b 2020

  7. [6]

    Kyriienko, A

    O. Kyriienko, A. E. Paine, and V. E. Elfving, Physical Review A, vol. 103, no. 5, 2021

    work page 2021

  8. [7]

    Quantum Computing in the NISQ era and beyond,

    J. Preskill, “Quantum Computing in the NISQ era and beyond,” Quantum, vol. 2, p. 79, Aug. 2018, issn: 2521-327X. doi: 10 . 22331 / q - 2018- 08- 06- 79. [Online]. Available: https: //doi.org/10.22331/q-2018-08-06-79

    doi:10.22331/q-2018-08-06-79 2018

Show all 27 references
  1. [8]

    Barren plateaus in quantum neural network training landscapes,

    J. R. McClean, S. Boixo, V. N. Smelyanskiy, R. Babbush, and H. Neven, “Barren plateaus in quantum neural network training landscapes,” Nature Communications, vol. 9, 2018. [Online]. Available: https : / / api . semanticscholar . org/CorpusID:4465524

    2018

  2. [9]

    Grant, L

    E. Grant, L. Wossnig, M. Ostaszewski, and M. Benedetti, Quantum, vol. 3, p. 214, 2019

    2019

  3. [10]

    Zhang, L

    K. Zhang, L. Liu, M.-H. Hsieh, and D. Tao, Escaping from the barren plateau via gaussian initializations in deep variational quantum cir- cuits, 2022. arXiv: 2203 . 09376 [quant-ph]. [Online]. Available: https://arxiv.org/abs/ 2203.09376

    2022 arXiv

  4. [11]

    Connecting ansatz expressibility to gra- dient magnitudes and barren plateaus,

    Z. Holmes, K. Sharma, M. Cerezo, and P. J. Coles, “Connecting ansatz expressibility to gra- dient magnitudes and barren plateaus,” PRX Quantum, vol. 3, no. 1, Jan. 2022, issn: 2691-

    2022

  5. [12]

    Gradmax: Growing neural networks using gradient infor- mation,

    U. Evci, M. Vladymyrov, T. Unterthiner, B. van Merrienboer, and F. Pedregosa, “Gradmax: Growing neural networks using gradient infor- mation,” ArXiv, vol. abs/2201.05125, 2022

    2022 arXiv

  6. [13]

    Duong, S

    T. Duong, S. T. Truong, M. Tam, B. Bach, J.-Y. Ryu, and J.-K. K. Rhee, Quantum neu- ral architecture search with quantum circuits metric and bayesian optimization, 2022. arXiv: 2206.14115 [quant-ph]

    2022 arXiv

  7. [14]

    Quantum circuit architecture search for variational quantum algorithms,

    Y. Du, T. Huang, S. You, M.-H. Hsieh, and D. Tao, “Quantum circuit architecture search for variational quantum algorithms,” npj Quantum Information, vol. 8, no. 1, May 2022,issn: 2056-

    2022

  8. [15]

    Hierarchical quan- tum circuit representations for neural architec- ture search,

    M. Lourens, I. Sinayskiy, D. K. Park, C. Blank, and F. Petruccione, “Hierarchical quan- tum circuit representations for neural architec- ture search,” npj Quantum Information, vol. 9, no. 1, pp. 1–15, 1 Aug. 5, 2023,issn: 2056-6387. doi: 10.1038/s41534-023-00747-z . 11

    2023 doi

  9. [16]

    Qubit-adapt-vqe: An adaptive algorithm for constructing hardware- efficient ans¨ atze on a quantum processor,

    H. L. Tang et al. , “Qubit-adapt-vqe: An adaptive algorithm for constructing hardware- efficient ans¨ atze on a quantum processor,”PRX Quantum, vol. 2, p. 020 310, 2 2021. doi: 10. 1103/PRXQuantum.2.020310 . [Online]. Avail- able: https://link.aps.org/doi/10.1103/ PRXQuantum.2.020310

    2021

  10. [17]

    Ef- fect of data encoding on the expressive power of variational quantum-machine-learning mod- els,

    M. Schuld, R. Sweke, and J. J. Meyer, “Ef- fect of data encoding on the expressive power of variational quantum-machine-learning mod- els,” Physical Review A , vol. 103, no. 3, Mar. 2021, issn: 2469-9934. doi: 10 . 1103 / physreva.103.032430

    2021

  11. [19]

    Robust data encod- ings for quantum classifiers,

    R. LaRose and B. Coyle, “Robust data encod- ings for quantum classifiers,” Phys. Rev. A, vol. 102, p. 032 420, 3 2020. doi: 10 . 1103 / PhysRevA . 102 . 032420. [Online]. Available: https : / / link . aps . org / doi / 10 . 1103 / PhysRevA.102.032420

    2020

  12. [20]

    Let quan- tum neural networks choose their own frequen- cies,

    B. Jaderberg, A. A. Gentile, Y. A. Berrada, E. Shishenina, and V. E. Elfving, “Let quan- tum neural networks choose their own frequen- cies,” Phys. Rev. A, vol. 109, p. 042 421, 4 2024. doi: 10 . 1103 / PhysRevA . 109 . 042421. [On- line]. Available: https://link.aps.org/doi/ ...

    2024 doi

  13. [21]

    An initialization strategy for ad- dressing barren plateaus in parametrized quan- tum circuits,

    E. Grant, L. Wossnig, M. Ostaszewski, and M. Benedetti, “An initialization strategy for ad- dressing barren plateaus in parametrized quan- tum circuits,” Quantum, vol. 3, p. 214, 2019, issn: 2521-327X. doi: 10.22331/q-2019-12- 09 - 214. [Online]. Available: https : / / doi . o...

    2019 doi

  14. [22]

    Available: http://dx.doi.org/ 10.1007/s42484-020-00036-4

    [Online]. Available: http://dx.doi.org/ 10.1007/s42484-020-00036-4

    doi

  15. [23]

    R. J. George, J. Zhao, J. Kossaifi, Z. Li, and A. Anandkumar, Incremental spatial and spectral learning of neural operators for solving large- scale pdes, 2024. arXiv: 2211.15188 [cs.LG]. [Online]. Available: https://arxiv.org/abs/ 2211.15188. 12 A Student Teacher The tables d...

    2024 arXiv

  16. [26]

    Fridovich-Keil, R

    S. Fridovich-Keil, R. Gontijo-Lopes, and R. Roelofs, Spectral bias in practice: The role of function frequency in generalization, 2022. arXiv: 2110 . 02424 [cs.LG]. [Online]. Avail- able: https://arxiv.org/abs/2110.02424

    2022 arXiv

  17. [2014]

    [Online]

    arXiv: 1411.4028 [quant-ph]. [Online]. Available: https : / / arxiv . org / abs / 1411 . 4028

    arXiv

  18. [3399]

    1103 / prxquantum

    doi: 10 . 1103 / prxquantum . 3 . 010313. [Online]. Available: http://dx.doi.org/10. 1103/PRXQuantum.3.010313

  19. [6387]

    1038 / s41534 - 022 - 00570 - y

    doi: 10 . 1038 / s41534 - 022 - 00570 - y. [Online]. Available: http://dx.doi.org/10. 1038/s41534-022-00570-y

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.