Pith. sign in

REVIEW 5 major objections 6 minor 75 references

Efficient Estimation and Sequential Optimization of Cost Functions in Variational Quantum Algorithms

T0 review · 5 major / 6 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read This paper claims that rewriting a parameterized quantum circuit as a weighted sum of unitaries turns cost-function optimization into per-parameter sweeps that beat COBYLA on nonlinear test problems.

desk verdict Rotosolve dressed in new applications; sound algebra, but the headline advantage over COBYLA is not yet established because the baseline is weak and the grid is unspecified. read the letter →

arxiv 2412.20972 v1 pith:Y2UE4T4D submitted 2024-12-30 quant-ph

Muhammad Umer , Eleftherios Mastorakis , Dimitris G. Angelakis This is my paper · ORCID
classification quant-ph
keywords variationalquantumalgorithmsparameterizedcircuitscostfunctionestimationsequentialoptimizationCOBYLAbaselineBurgersequationnonlinearSchrödingerunitarydecomposition
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 tries to establish that the classical optimization step inside variational quantum algorithms can be made cheaper and more reliable by changing how the cost function is represented. The key move is to expand the parameterized circuit as a weighted sum of fixed unitary operators, so that cost functions become closed trigonometric expressions in each variational angle. That turns per-parameter optimization into a one-dimensional curve search that can be carried out classically after a small number of quantum measurements, and it makes derivatives of any order available for free. The authors package this into a coordinate-descent optimizer called sequential grid-based explicit optimization (SGEO) and test it on the one-dimensional Burgers equation and the nonlinear Schrödinger equation. On those tests SGEO reaches variational states with infidelity below roughly one to two percent while the COBYLA optimizer visibly lags, which is the claim a sympathetic reader would take away.

What carries the argument

The central object is the two-term unitary decomposition of a single-qubit rotation, $e^{-i\lambda_j P/2} = \cos(\lambda_j/2) \hat{I} - i \sin(\lambda_j/2) \hat{P}$, which expands the whole parameterized circuit into a weighted sum of fixed unitaries with trigonometric coefficients. Inserting this sum into the two cost functions gives closed-form expressions for the cost along each parameter axis, with all nonlocal information packed into coefficients that are estimated by a fixed small number of quantum circuits. The optimization protocol SGEO is coordinate descent on top of those curves: for each parameter in turn it evaluates the entire one-dimensional cost function classically, updates the parameter to its exact minimizer, and repeats for a fixed number of sweeps.

What would settle it

Construct a cost landscape with a strict saddle whose per-coordinate slices all have a minimum at the saddle, and run SGEO on it: if it terminates at the saddle while a gradient-based optimizer moves to a lower point, the claim that SGEO consistently outperforms COBYLA is refuted. Alternatively, scale the Burgers or NLSE problem to more qubits and record the number of sweeps needed to reach a fixed infidelity; rapid growth of that sweep count would erase the reported advantage.

Watch

Extended reading notes

Core claim

The central claim is that a generic parameterized quantum circuit, built from single-qubit rotations and fixed entangling gates, can be rewritten as a weighted sum of fixed unitaries: each rotation $e^{-i\lambda_j P/2}$ splits into $\cos(\lambda_j/2)$ times the identity and $\sin(\lambda_j/2)$ times a fixed Pauli unitary, and expanding every rotation expresses the whole circuit as an explicitly parametrized linear combination of $2^m$ unitaries. Substituting that expansion into the squared-residual cost function (Burgers dynamics) or the expectation-value cost function (ground-state problems) yields closed-form functions of each $\lambda_j$ whose coefficients, such as $\alpha_0/\alpha_\pi$ or $\kappa_{0,0}/\kappa_{\pi,\pi}$, are measured by a small constant number of quantum circuits. Once those coefficients are known, the full cost curve over $\lambda_j \in [-\pi, \pi)$ and its arbitrary derivatives are evaluated classically, so each coordinate update can jump directly to the exact one-dimensional minimum. The paper reports that this sequential sweep, SGEO, consistently obtains lower cost and higher state fidelity than COBYLA in the Burgers and nonlinear Schrödinger applications, and that in the investigated regimes the variational states remain within roughly one to two percent infidelity of the classical target states.

Load-bearing premise

The load-bearing premise is that a fixed, hand-chosen number of coordinate-descent sweeps (five or ten in the numerical tests) is enough to reach the global minimum of the cost function; the paper does not prove convergence, so if the needed sweeps grow with problem size the reported advantage over COBYLA would disappear.

Editorial extensions

If this is right

  • Any derivative of the squared-residual or expectation cost function becomes classically computable once the fixed-unitary overlaps are measured, so parameter-shift evaluations are not needed for the reported tests.
  • For the Burgers equation test cases, including turbulent shock formation, SGEO keeps the infidelity below roughly one to two percent over the simulated dynamics, with laminar-regime fidelities above 99 percent.
  • For the nonlinear Schrödinger ground-state problem, SGEO reaches the minimum energy across weak, intermediate, and strong nonlinearity, while COBYLA stalls in all three regimes.
  • Because each update uses the full parameter domain rather than a local step, the optimizer is gradient-free and does not require step-size tuning.
  • The decomposition extends to parameterized two-qubit gates such as (i)SWAP, so the same strategy applies beyond single-qubit rotations.

Reading between the lines

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

  • If the one-dimensional slices of the cost landscape are as informative as these tests suggest, the same decomposition could be used to diagnose flat regions and barren plateaus before optimization, by checking how flat the cost curves are across parameters.
  • The reported comparison is not resource-neutral in circuit count, since SGEO pays a small constant factor for the extra overlap measurements per sweep; a fair head-to-head that fixes total circuit evaluations at larger qubit numbers would sharpen where the advantage holds.
  • The closed-form derivatives also open a natural hybrid route: use SGEO sweeps to escape local traps, then switch to gradient-based refinement once near the minimum.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

5 major / 6 minor

Summary. The paper proposes a sequential optimization method (SGEO) for variational quantum algorithms, based on expressing the parameterized quantum circuit as a linear combination of unitaries evaluated at parameter values 0 and π. This enables the analytic construction of the cost function over the full range of a single parameter, and hence exact or grid-based line searches and arbitrary derivatives without additional quantum resources. The method is applied to two problems: the viscous Burgers equation (squared residual cost) and the ground state of the one-dimensional nonlinear Schrödinger equation (expectation cost). The paper claims that SGEO consistently outperforms COBYLA in convergence speed and accuracy. The algebraic derivations in Sec. II are internally consistent, but the numerical evidence is incomplete: the grid resolution is unspecified, the COBYLA baseline is configured with a small initial simplex and few iterations, and no statistical uncertainties are reported.

Significance. If substantiated, the algorithm would be a useful addition to the VQA optimization toolbox, particularly because for each parameter it obtains the full one-dimensional cost landscape from two or three measured quantities and can compute arbitrary derivatives classically. The paper also provides explicit circuit constructions for the overlap and expectation-value measurements, and the formulas in Sec. II are derived cleanly. However, the paper does not yet establish the central comparative claim against COBYLA; the single-parameter line search is closely related to Rotosolve/sequential minimal optimization, so the novelty lies mainly in the resource-efficient estimation formulas and their application to nonlinear physics problems. With a properly specified grid resolution and a fairer baseline comparison, the claimed advantage could be convincingly demonstrated. As it stands, the evidence is suggestive but not conclusive.

major comments (5)
  1. [Sec. II.D / Algorithm 1] The classical line search in Algorithm 1 is described as 'grid-based' (the algorithm's name is SGEO), but the number of grid points or the resolution is never specified. The text in Sec. II.D states that the cost is 'evaluated for λj ∈ [−π, π) on a classical computer' and that the minimizing λj is found, yet no discretization is reported in either application. This omission affects both reproducibility and the resource comparison: if the grid is coarse, the reported infidelities are not guaranteed to reflect the true line-search minima, and if the grid is fine, the classical optimization cost is hidden. Please specify the grid density or the convergence tolerance used, and include the classical line-search cost in the complexity accounting.
  2. [Sec. II.E and Secs. III-IV] The COBYLA baseline is configured with rhobeg = π/16 and tol = 10^-10, and is given 100–300 iterations. With a parameter domain of width 2π, this choice of rhobeg gives an initial simplex that is much smaller than the default (rhobeg = 1.0 in standard implementations), which can slow the trust-region search. The paper does not investigate the sensitivity of its conclusions to this hyperparameter. A fair comparison would report COBYLA results for a range of rhobeg values (e.g., the default, 0.5, 0.1) and larger iteration budgets, or provide a clear justification for why the chosen value is representative.
  3. [Figs. 3-6] All reported infidelity and energy curves are from single runs with no error bars. Given that each circuit evaluation uses 5×10^4 shots, the overlap and expectation-value estimates carry shot noise, and the text itself notes fluctuations due to shot noise (e.g., Sec. III, discussion of Fig. 4). The claim that SGEO 'consistently outperforms' COBYLA is a statistical claim, and it is not supported by the present data. Please provide repeated runs over multiple random initializations and report means with confidence intervals (or at least standard errors), and avoid claiming 'consistent' superiority without this evidence.
  4. [Sec. III vs. Sec. IV] The computational budget comparison is not made on the same footing in the Burgers section. In Sec. III, the iteration counts are N_COBYLA = 100/200 and NSGEO = 5/10, which give 100–200 COBYLA parameter updates but m×5 to m×10 SGEO parameter updates (e.g., 80–200 for the 4-qubit system). The paper should plot the cost and infidelity against the number of circuit evaluations, as is done in Sec. IV for the NLSE, so that the reader can fairly assess the quantum-resource cost of each method. The current Fig. 4 plots against 'parameter updates,' which does not account for the different number of circuit evaluations per update.
  5. [Sec. II.D and Sec. IV] The text claims that SGEO 'effectively identifies the global optimal point' (Sec. IV) and that the method avoids local minima and barren plateaus (Sec. V). However, Algorithm 1 is a coordinate-descent procedure with a fixed number of sweeps, and no convergence proof is provided. In general, coordinate descent in nonconvex landscapes can converge to local minima; the paper's evidence for global convergence is limited to the specific test instances. Please either limit the claims to the tested problems or provide a theoretical analysis for the class of cost functions considered (e.g., showing that the line-search updates converge to a stationary point under the given noise model).
minor comments (6)
  1. [Sec. II.D] In the first paragraph of Sec. II.D, 'SEGO' is a typo; it should read 'SGEO'.
  2. [Sec. IV] The sentence 'Eqs. (15 - 17) require fourteen such circuits' appears to omit the kinetic-energy term, which is in Eq. (18). Please correct the equation numbering or clarify which set of equations requires fourteen circuits.
  3. [Sec. IV] The phrase 'In contrast to Eq. (14), which requires only four quantum circuits' is confusing because Eq. (14) is the NLSE itself, not a cost function. It seems the authors mean the cost components in Eq. (15); please reword.
  4. [Sec. II.A] The notation 'λj0' and the superscripts 0 and π on the unitaries are used before they are explicitly defined; please introduce the convention that subscripts/superscripts 0 and π denote fixing the corresponding parameter to those values.
  5. [Algorithm 1] The caption 'Algorithm 1:Sequential Grid-Based Explicit Optimization' lacks a space after the colon; please format as 'Algorithm 1: Sequential Grid-Based Explicit Optimization'.
  6. [Sec. II.E] The text writes '5 × 104 shots'; please render the exponent properly as '5 × 10^4 shots'.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the cost-function coefficients are device-measured, the 1D cost curves follow from the PQC linear-combination algebra, and the single self-citation [23] is a non-load-bearing building block.

full rationale

The derivation chain is self-contained. The central coefficients α, β, κ, ζ and the G terms are defined as overlaps or expectation values that are measured on the quantum device (Secs. II.B, II.C, III, IV), and the 1D cost expressions (Eqs. 6, 9, 13, 16-18) are exact algebraic consequences of the PQC decomposition in Eq. (4), not fitted or target-dependent forms. SGEO (Algorithm 1) minimizes those exact 1D curves along each coordinate, and the reported infidelities are evaluated against independent classical references (the classical Burgers solution and the imaginary-time-evolution NLSE ground state), not against the cost function itself. No parameter is fitted to the claimed outputs, and no prediction reduces by construction to an input. The only self-citation in the derivation path is Ref. [23], used for the NLSE energy decomposition; that is a building block shared with the independent Refs. [21,22], it does not assert uniqueness or forbid alternatives, and the central SGEO-versus-COBYLA claim stands or falls on the simulations, not on Ref. [23]. Benchmarking concerns about the COBYLA baseline (rhobeg = π/16, fixed iteration counts, single runs, unspecified grid resolution) are correctness and experimental-design issues, not circularity.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

The central derivation relies on standard trigonometric identities and on measured hardware coefficients. The main non-standard assumptions are the expressivity of the fixed-depth ansatz and the sufficiency of a small fixed number of coordinate-descent sweeps; neither is proven.

free parameters (4)
  • Number of ansatz layers d = d=2,3,4 depending on regime and qubit count
    Chosen by hand per problem (Sec. III, IV); affects expressivity and optimization difficulty; no selection criterion is given.
  • Iteration counts N_SGEO and N_COBYLA = N_SGEO=5/10, N_COBYLA=100/200/300
    Set per regime; determines the circuit-evaluation budget and directly shapes the comparison (Sec. III, IV).
  • Grid resolution for SGEO line search = not stated
    Algorithm 1 says evaluate cost over λ_j ∈ [−π,π) and find the minimum, but resolution is unspecified; needed for reproducibility.
  • COBYLA hyperparameters = rhobeg=π/16, tol=10^-10
    Standard but arbitrary; may disadvantage COBYLA (Sec. II.E).
assumptions (5)
  • standard math Euler expansion of single-qubit rotations
    Eq. (2) uses U(λ)=cos(λ/2)I - i sin(λ/2)P; universally valid.
  • domain assumption Real-amplitude ansatz expressivity
    Sec. II.E assumes the chosen depth d is sufficient to represent the target states; the paper does not assess expressivity.
  • domain assumption Finite-difference/Euler discretization accuracy
    Sec. III discretizes Burgers with δx and τ=δx/10; discretization error is not quantified.
  • domain assumption Finite-shot estimates represent exact expectation values
    Coefficients α/κ/γ are estimated with 5×10^4 shots; statistical noise is acknowledged but not characterized (Sec. III, Fig. 4 caption).
  • ad hoc to paper Convergence of coordinate descent with fixed sweeps
    Algorithm 1 runs a fixed number of sweeps; no convergence proof, and the numbers 5/10 are chosen per problem.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Efficient Estimation and Sequential Optimization of Cost Functions in Variational Quantum Algorithms." pith.science (2026). https://pith.science/paper/Y2UE4T4D

@misc{pith2026241220972,
  author       = {Pith},
  title        = {Pith review of: Efficient Estimation and Sequential Optimization of Cost Functions in Variational Quantum Algorithms},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/Y2UE4T4D}},
  note         = {Machine review of arXiv:2412.20972}
}
read the original abstract

Classical optimization is a cornerstone of the success of variational quantum algorithms, which often require determining the derivatives of the cost function relative to variational parameters. The computation of the cost function and its derivatives, coupled with their effective utilization, facilitates faster convergence by enabling smooth navigation through complex landscapes, ensuring the algorithm's success in addressing challenging variational problems. In this work, we introduce a novel optimization methodology that conceptualizes the parameterized quantum circuit as a weighted sum of distinct unitary operators, enabling the cost function to be expressed as a sum of multiple terms. This representation facilitates the efficient evaluation of nonlocal characteristics of cost functions, as well as their arbitrary derivatives. The optimization protocol then utilizes the nonlocal information on the cost function to facilitate a more efficient navigation process, ultimately enhancing the performance in the pursuit of optimal solutions. We utilize this methodology for two distinct cost functions. The first is the squared residual of the variational state relative to a target state, which is subsequently employed to examine the nonlinear dynamics of fluid configurations governed by the one-dimensional Burgers' equation. The second cost function is the expectation value of an observable, which is later utilized to approximate the ground state of the nonlinear Schr\"{o}dinger equation. Our findings reveal substantial enhancements in convergence speed and accuracy relative to traditional optimization methods, even within complex, high-dimensional landscapes. Our work contributes to the advancement of optimization strategies for variational quantum algorithms, establishing a robust framework for addressing a range of computationally intensive problems across numerous applications.

Figures

Figures reproduced from arXiv: 2412.20972 by the authors.

Figure 1
Figure 1. FIG. 1. Structure of the real-amplitude ansatz. Here, [PITH_FULL_IMAGE:figures/full_fig_p006_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2. Design of quantum circuits for evaluating vari [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗
Figure 3
Figure 3. FIG. 3. Nonlinear dynamics of fluid configurations governed by the Burgers [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: indicate the minimum cost function value achieved by SGEO, illustrating that the SGEO protocol can ter￾minate before exhausting all iterations. Future research could focus on developing a mechanism to terminate the SGEO protocol upon reaching the cost function mini￾mum…
Figure 6
Figure 6. Figure 6: FIG. 6. Analysis of the convergence of energy difference (∆ [PITH_FULL_IMAGE:figures/full_fig_p012_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

75 extracted references · 43 canonical work pages

  1. [1]

    P. W. Shor, Algorithms for quantum computation: dis- crete logarithms and factoring, Proceedings 35th Annual Symposium on Foundations of Computer Science , 124 (1994)

    work page 1994

  2. [2]

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

    work page 1996

  3. [3]

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

    2009

  4. [4]

    D. S. Abrams and S. Lloyd, Simulation of many-body fermi systems on a universal quantum computer, Phys. Rev. Lett. 79, 2586 (1997); Quantum algorithm provid- ing exponential speed increase for finding eigenvalues and eigenvectors, Phys. Rev. Lett. 83, 5162 (1999)

    work page 1997

  5. [5]

    Khodjasteh and D

    K. Khodjasteh and D. A. Lidar, Fault-tolerant quan- tum dynamical decoupling, Phys. Rev. Lett. 95, 180501 (2005)

    work page 2005

  6. [6]

    Temme, S

    K. Temme, S. Bravyi, and J. M. Gambetta, Error mitiga- tion for short-depth quantum circuits, Phys. Rev. Lett. 119, 180509 (2017)

    2017

  7. [7]

    S. Endo, S. C. Benjamin, and Y. Li, Practical quantum error mitigation for near-future applications, Phys. Rev. X 8, 031027 (2018)

    2018

  8. [8]

    Y. Kim, A. Eddins, S. Anand, K. X. Wei, E. Van Den Berg, S. Rosenblatt, H. Nayfeh, Y. Wu, M. Zale- tel, K. Temme, et al., Evidence for the utility of quan- tum computing before fault tolerance, Nature 618, 500 (2023)

    work page 2023

Show all 75 references
  1. [9]

    Cerezo, A

    M. Cerezo, A. Arrasmith, R. Babbush, S. C. Benjamin, S. Endo, K. Fujii, J. R. McClean, K. Mitarai, X. Yuan, L. Cincio, et al., Variational quantum algorithms, Nature Reviews Physics 3, 625 (2021)

    2021

  2. [10]

    Bharti, A

    K. Bharti, A. Cervera-Lierta, T. H. Kyaw, T. Haug, S. Alperin-Lea, A. Anand, M. Degroote, H. Heimonen, J. S. Kottmann, T. Menke, W.-K. Mok, S. Sim, L.-C. Kwek, and A. Aspuru-Guzik, Noisy intermediate-scale quantum algorithms, Rev. Mod. Phys.94, 015004 (2022)

    2022

  3. [11]

    Peruzzo, J

    A. Peruzzo, J. McClean, P. Shadbolt, M.-H. Yung, X.-Q. Zhou, P. J. Love, A. Aspuru-Guzik, and J. L. O’brien, A variational eigenvalue solver on a photonic quantum processor, Nature communications 5, 4213 (2014)

    2014

  4. [12]

    Wecker, M

    D. Wecker, M. B. Hastings, and M. Troyer, Progress to- wards practical quantum variational algorithms, Phys. Rev. A 92, 042303 (2015)

    2015

  5. [13]

    P. J. J. O’Malley, R. Babbush, I. D. Kivlichan, J. Romero, J. R. McClean, R. Barends, J. Kelly, P. Roushan, A. Tranter, N. Ding, B. Campbell, Y. Chen, Z. Chen, B. Chiaro, A. Dunsworth, A. G. Fowler, E. Jef- frey, E. Lucero, A. Megrant, J. Y. Mutus, M. Neeley, C. Neill, C. Quin...

    2016

  6. [14]

    Kandala, A

    A. Kandala, A. Mezzacapo, K. Temme, M. Takita, M. Brink, J. M. Chow, and J. M. Gambetta, Hardware- efficient variational quantum eigensolver for small molecules and quantum magnets, Nature 549, 242 (2017)

    2017

  7. [15]

    Hempel, C

    C. Hempel, C. Maier, J. Romero, J. McClean, T. Monz, H. Shen, P. Jurcevic, B. P. Lanyon, P. Love, R. Babbush, A. Aspuru-Guzik, R. Blatt, and C. F. Roos, Quantum chemistry calculations on a trapped-ion quantum simu- lator, Phys. Rev. X 8, 031022 (2018)

    2018

  8. [16]

    Ganzhorn, D

    M. Ganzhorn, D. Egger, P. Barkoutsos, P. Ollitrault, G. Salis, N. Moll, M. Roth, A. Fuhrer, P. Mueller, S. Wo- 15 erner, I. Tavernelli, and S. Filipp, Gate-efficient simula- tion of molecular eigenstates on a quantum computer, Phys. Rev. Appl. 11, 044092 (2019)

    2019

  9. [17]

    G. A. Quantum, Collaborators* †, F. Arute, K. Arya, R. Babbush, D. Bacon, J. C. Bardin, R. Barends, S. Boixo, M. Broughton, B. B. Buckley, et al., Hartree- fock on a superconducting qubit quantum computer, Sci- ence 369, 1084 (2020)

    2020

  10. [18]

    T. N. N. Cong and H. L. Thi, Variational quantum algo- rithms in finance, in Proceedings of Ninth International Congress on Information and Communication Technol- ogy, edited by X.-S. Yang, S. Sherratt, N. Dey, and A. Joshi (Springer Nature Singapore, Singapore, 2024) pp. 15–25

    2024

  11. [19]

    E. X. Huber, B. Y. Tan, P. R. Griffin, and D. G. An- gelakis, Exponential qubit reduction in optimization for financial transaction settlement, EPJ Quantum Technol- ogy 11, 52 (2024)

    2024

  12. [20]

    Sarma, T

    A. Sarma, T. W. Watts, M. Moosa, Y. Liu, and P. L. McMahon, Quantum variational solving of nonlinear and multidimensional partial differential equations, Phys. Rev. A 109, 062616 (2024)

    2024

  13. [21]

    Lubasch, J

    M. Lubasch, J. Joo, P. Moinier, M. Kiffner, and D. Jaksch, Variational quantum algorithms for nonlinear problems, Physical Review A 101, 010301 (2020)

    2020

  14. [22]

    Jaksch, P

    D. Jaksch, P. Givi, A. J. Daley, and T. Rung, Variational quantum algorithms for computational fluid dynamics, AIAA journal 61, 1885 (2023)

    2023

  15. [23]

    M. Umer, E. Mastorakis, S. Evangelou, and D. G. Angelakis, Probing the limits of varia- tional quantum algorithms for nonlinear ground states on real quantum hardware: The ef- fects of noise, arXiv preprint arXiv:2403.16426 https://doi.org/10.48550/arXiv.2403.16426 (2024)

    2024 doi arXiv

  16. [24]

    Cirstoiu, Z

    C. Cirstoiu, Z. Holmes, J. Iosue, L. Cincio, P. J. Coles, and A. Sornborger, Variational fast forwarding for quan- tum simulation beyond the coherence time, npj Quantum Information 6, 82 (2020)

    2020

  17. [25]

    S.-H. Lin, R. Dilip, A. G. Green, A. Smith, and F. Poll- mann, Real- and imaginary-time evolution with com- pressed quantum circuits, PRX Quantum 2, 010342 (2021)

    2021

  18. [26]

    Linteau, S

    D. Linteau, S. Barison, N. H. Lindner, and G. Carleo, Adaptive projected variational quantum dynamics, Phys. Rev. Res. 6, 023130 (2024)

    2024

  19. [27]

    Farhi, J

    E. Farhi, J. Goldstone, and S. Gutmann, A quantum approximate optimization al- gorithm, arXiv preprint arXiv:1411.4028 https://doi.org/10.48550/arXiv.1411.4028 (2014), https://arxiv.org/abs/1411.4028

    2014 doi arXiv

  20. [28]

    Pagano, A

    G. Pagano, A. Bapat, P. Becker, K. S. Collins, A. De, P. W. Hess, H. B. Kaplan, A. Kyprianidis, W. L. Tan, C. Baldwin, et al., Quantum approximate optimization of the long-range ising model with a trapped-ion quan- tum simulator, Proceedings of the National Academy of Sciences...

    2020

  21. [29]

    Tan, M.-A

    B. Tan, M.-A. Lemonde, S. Thanasilp, J. Tangpanitanon, and D. G. Angelakis, Qubit-efficient encoding schemes for binary optimisation problems, Quantum 5, 454 (2021)

    2021

  22. [30]

    L. Zhu, H. L. Tang, G. S. Barron, F. A. Calderon-Vargas, N. J. Mayhall, E. Barnes, and S. E. Economou, Adaptive quantum approximate optimization algorithm for solving combinatorial problems on a quantum computer, Phys. Rev. Res. 4, 033029 (2022)

    2022

  23. [31]

    Benedetti, E

    M. Benedetti, E. Lloyd, S. Sack, and M. Fiorentini, Pa- rameterized quantum circuits as machine learning mod- els, Quantum Science and Technology 4, 043001 (2019)

    2019

  24. [32]

    D. Zhu, N. M. Linke, M. Benedetti, K. A. Landsman, N. H. Nguyen, C. H. Alderete, A. Perdomo-Ortiz, N. Ko- rda, A. Garfoot, C. Brecque, et al., Training of quantum circuits on a hybrid quantum computer, Science advances 5, eaaw9918 (2019)

    2019

  25. [33]

    Tangpanitanon, S

    J. Tangpanitanon, S. Thanasilp, N. Dangniam, M.-A. Lemonde, and D. G. Angelakis, Expressibility and train- ability of parametrized analog quantum systems for ma- chine learning applications, Phys. Rev. Res. 2, 043364 (2020)

    2020

  26. [34]

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

    2016

  27. [35]

    J. R. McClean, S. Boixo, V. N. Smelyanskiy, R. Bab- bush, and H. Neven, Barren plateaus in quantum neural network training landscapes, Nature communications 9, 4812 (2018)

    2018

  28. [36]

    Bittel and M

    L. Bittel and M. Kliesch, Training variational quantum algorithms is np-hard, Phys. Rev. Lett. 127, 120502 (2021)

    2021

  29. [37]

    Fontana, M

    E. Fontana, M. Cerezo, A. Arrasmith, I. Rungger, and P. J. Coles, Non-trivial symmetries in quantum land- scapes and their resilience to quantum noise, Quantum 6, 804 (2022)

    2022

  30. [38]

    E. R. Anschuetz and B. T. Kiani, Quantum variational algorithms are swamped with traps, Nature Communica- tions 13, 7760 (2022)

    2022

  31. [39]

    Larocca, P

    M. Larocca, P. Czarnik, K. Sharma, G. Muraleed- haran, P. J. Coles, and M. Cerezo, Diagnosing barren plateaus with tools from quantum optimal control, Quantum 6, 824 (2022); M. Larocca, S. Thanasilp, S. Wang, K. Sharma, J. Biamonte, P. J. Coles, L. Cincio, J. R. McClean, Z. Ho...

    2022 doi arXiv

  32. [40]

    H. Qi, L. Wang, H. Zhu, A. Gani, and C. Gong, The bar- ren plateaus of quantum neural networks: review, tax- onomy and trends, Quantum Information Processing 22, 435 (2023)

    2023

  33. [41]

    C. G. Broyden, The Convergence of a Class of Double- rank Minimization Algorithms 1. General Considera- tions, IMA Journal of Applied Mathematics 6, 76 (1970); R. Fletcher, A new approach to variable metric algo- rithms, The Computer Journal 13, 317 (1970); D. Gold- farb, A fa...

    1970

  34. [42]

    Kraft, A software package for sequential quadratic programming, Forschungsbericht- Deutsche Forschungs- und Versuchsanstalt fur Luft- und Raumfahrt (1988)

    D. Kraft, A software package for sequential quadratic programming, Forschungsbericht- Deutsche Forschungs- und Versuchsanstalt fur Luft- und Raumfahrt (1988)

    1988

  35. [43]

    D. P. Kingma, Adam: A method for stochas- tic optimization, arXiv preprint arXiv:1412.6980 https://doi.org/10.48550/arXiv.1412.6980 (2014)

    2014 doi arXiv

  36. [44]

    M. J. Powell, A direct search optimization method that models the objective and constraint functions by lin- ear interpolation (Springer, 1994); Direct search algo- rithms for optimization calculations, Acta numerica 7, 16 287 (1998); A view of algorithms for optimization with...

    1998

  37. [45]

    J. C. Spall, A stochastic approximation technique for generating maximum likelihood parameter estimates, in 1987 American control conference (IEEE, 1987) pp. 1161–1167; Multivariate stochastic approximation us- ing a simultaneous perturbation gradient approximation, IEEE trans...

    1992

  38. [46]

    J. A. Nelder and R. Mead, A simplex method for function minimization, The computer journal 7, 308 (1965)

    1965

  39. [47]

    Schuld, V

    M. Schuld, V. Bergholm, C. Gogolin, J. Izaac, and N. Kil- loran, Evaluating analytic gradients on quantum hard- ware, Phys. Rev. A 99, 032331 (2019)

    2019

  40. [48]

    G. E. Crooks, Gradients of parameterized quan- tum gates using the parameter-shift rule and gate decomposition, arXiv preprint arXiv:1905.13311 https://doi.org/10.48550/arXiv.1905.13311 (2019)

    2019 doi arXiv

  41. [49]

    A. Mari, T. R. Bromley, and N. Killoran, Estimating the gradient and higher-order derivatives on quantum hard- ware, Phys. Rev. A 103, 012405 (2021)

    2021

  42. [50]

    Banchi and G

    L. Banchi and G. E. Crooks, Measuring analytic gradi- ents of general quantum evolution with the stochastic parameter shift rule, Quantum 5, 386 (2021)

    2021

  43. [51]

    Hubregtsen, F

    T. Hubregtsen, F. Wilde, S. Qasim, and J. Eisert, Single- component gradient rules for variational quantum al- gorithms, Quantum Science and Technology 7, 035008 (2022)

    2022

  44. [52]

    Mitarai and K

    K. Mitarai and K. Fujii, Methodology for replacing indi- rect measurements with direct measurements, Phys. Rev. Res. 1, 013006 (2019)

    2019

  45. [53]

    A. W. Harrow and J. C. Napp, Low-depth gradient mea- surements can improve convergence in variational hy- brid quantum-classical algorithms, Phys. Rev. Lett. 126, 140502 (2021)

    2021

  46. [54]

    Virtanen, R

    P. Virtanen, R. Gommers, T. E. Oliphant, M. Haber- land, T. Reddy, D. Cournapeau, E. Burovski, P. Peter- son, W. Weckesser, J. Bright,et al., Scipy 1.0: fundamen- tal algorithms for scientific computing in python, Nature methods 17, 261 (2020)

    2020

  47. [55]

    Ostaszewski, E

    M. Ostaszewski, E. Grant, and M. Benedetti, Structure optimization for parameterized quantum circuits, Quan- tum 5, 391 (2021)

    2021

  48. [56]

    R. M. Parrish, J. T. Iosue, A. Ozaeta, and P. L. McMa- hon, A jacobi diagonalization and anderson accelera- tion algorithm for variational quantum algorithm pa- rameter optimization, arXiv preprint arXiv:1904.03206 https://doi.org/10.48550/arXiv.1904.03206 (2019)

    2019 doi arXiv

  49. [57]

    K. M. Nakanishi, K. Fujii, and S. Todo, Sequential minimal optimization for quantum-classical hybrid algo- rithms, Phys. Rev. Res. 2, 043158 (2020)

    2020

  50. [58]

    M. A. Nielsen and I. L. Chuang, Quantum computation and quantum information (Cambridge University Press, 2010)

    2010

  51. [59]

    Barenco, C

    A. Barenco, C. H. Bennett, R. Cleve, D. P. DiVincenzo, N. Margolus, P. Shor, T. Sleator, J. A. Smolin, and H. Weinfurter, Elementary gates for quantum computa- tion, Phys. Rev. A 52, 3457 (1995)

    1995

  52. [60]

    M. S. Anis, H. Abraham, R. A. AduOffei, G. Agliardi, M. Aharoni, I. Y. Akhalwaya, G. Aleksandrowicz, T. Alexander, M. Amy, S. Anagolum, et al., Qiskit: An open-source framework for quantum computing, Qiskit/qiskit 10.5281/zenodo.2573505 (2021)

    2021 doi

  53. [61]

    Bec and K

    J. Bec and K. Khanin, Burgers turbulence, Physics re- ports 447, 1 (2007)

    2007

  54. [62]

    (black crosses and orange diamonds) are observed at t = 0 .9. In contrast, the variational state obtained from the VQA using the COBYLA optimizer (blue cir- cles) deviates significantly from the target classical state, resulting in a distinctly different state at t = 0.9. The ...

  55. [63]

    Bendaas, Periodic wave shock solutions of burgers equations, Cogent Mathematics & Statistics 5, 1463597 (2018)

    S. Bendaas, Periodic wave shock solutions of burgers equations, Cogent Mathematics & Statistics 5, 1463597 (2018)

    2018

  56. [64]

    G. P. Agrawal, Nonlinear fiber optics, in Nonlinear Sci- ence at the Dawn of the 21st Century, edited by P. L. Christiansen, M. P. Sørensen, and A. C. Scott (Springer Berlin Heidelberg, Berlin, Heidelberg, 2000) pp. 195–211

    2000

  57. [65]

    Nakkeeran, Bright and dark optical solitons in fiber media with higher-order effects, Chaos, Solitons & Frac- tals 13, 673 (2002)

    K. Nakkeeran, Bright and dark optical solitons in fiber media with higher-order effects, Chaos, Solitons & Frac- tals 13, 673 (2002)

    2002

  58. [66]

    Triki, C

    H. Triki, C. Bensalem, A. Biswas, S. Khan, Q. Zhou, S. Adesanya, S. P. Moshokoa, and M. Belic, Self-similar optical solitons with continuous-wave background in a quadratic–cubic non-centrosymmetric waveguide, Optics Communications 437, 392 (2019)

    2019

  59. [67]

    Sulem and P.-L

    C. Sulem and P.-L. Sulem, The nonlinear schr¨ odinger equation: self-focusing and wave collapse (Springer Sci- ence & Business Media, 1999)

    1999

  60. [68]

    E. P. Gross, Structure of a quantized vortex in boson systems, Il Nuovo Cimento (1955-1965) 20, 454 (1961)

    1961

  61. [69]

    L. P. Pitaevskii, Vortex lines in an imperfect bose gas, Sov. Phys. JETP 13, 451 (1961)

    1961

  62. [70]

    Dalfovo, S

    F. Dalfovo, S. Giorgini, L. P. Pitaevskii, and S. Stringari, Theory of bose-einstein condensation in trapped gases, Reviews of modern physics 71, 463 (1999)

    1999

  63. [71]

    A. J. Leggett, Bose-einstein condensation in the alkali gases: Some fundamental concepts, Reviews of modern physics 73, 307 (2001)

    2001

  64. [72]

    Pitaevskii and S

    L. Pitaevskii and S. Stringari, Bose-Einstein condensa- tion (2003)

    2003

  65. [73]

    Edwards and K

    M. Edwards and K. Burnett, Numerical solution of the nonlinear schr¨ odinger equation for small samples of trapped neutral atoms, Physical Review A 51, 1382 (1995)

    1995

  66. [74]

    Lubasch, P

    M. Lubasch, P. Moinier, and D. Jaksch, Multigrid renor- malization, Journal of Computational Physics 372, 587 (2018)

    2018

  67. [75]

    Vedral, A

    V. Vedral, A. Barenco, and A. Ekert, Quantum networks for elementary arithmetic operations, Physical Review A 54, 147 (1996)

    1996

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.