Pith. sign in

REVIEW 5 major objections 7 minor 52 references

Non-Variational Quantum Random Access Optimization with Alternating Operator Ansatz

T0 review · 5 major / 7 minor · reviewed 2026-08-08 · deepseek-v4-flash

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

Pith's one-line read This paper proposes QAOA-for-QRAO, a non-variational scheme that runs quantum random access optimization with one fixed set of instance-independent parameters and needs up to three times fewer qubits than standard QAOA.

desk verdict A useful empirical study of QAOA on QRAO's relaxed Hamiltonian, with fixed parameters that appear to transfer within one graph family; the main soft spots are undisclosed fitted angles and a confounded mixer comparison. read the letter →

arxiv 2502.04277 v2 pith:7BPHYWA4 submitted 2025-02-06 quant-ph cs.ET

Zichang He , Rudy Raymond , Ruslan Shaydulin , Marco Pistoia This is my paper · ORCID
classification quant-phcs.ET MSC 81P6868Q12 PACS 03.67.Ac03.67.Lx
keywords quantumrandomaccessoptimizationQAOAnon-variationalfixedparametersMaxCutcodesTrotterizationentanglemententropy
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

Quantum Random Access Optimization (QRAO) encodes several optimization variables into each qubit, but every implementation so far required a variational loop to tune parameters for each instance. This paper proposes running QRAO with the Quantum Alternating Operator Ansatz (QAOA) using a single fixed set of instance-independent parameters, and benchmarks the idea on MaxCut. It reports that on random 3-regular graphs up to 26 nodes and depth 6, the fixed parameters match per-instance optimized parameters in approximation quality, and that only a couple of Trotter steps are needed to implement the non-diagonal cost layer. If the results hold, space-efficient quantum optimization could run on early fault-tolerant hardware without the overhead of variational training.

What carries the argument

The central object is the relaxed, non-diagonal Hamiltonian \(\tilde{H}_C\) built from a (3,1)-quantum random access code, which encodes three binary variables into one qubit through Pauli X, Y, and Z assignments. The paper drives this Hamiltonian with alternating cost and mixer unitaries, selecting the Z mixer with the |0⟩ initial state rather than the usual X mixer. Because the Pauli terms in \(\tilde{H}_C\) do not commute, the cost layer is implemented by first-order Trotter or Grouped Trotter decomposition, with only a few Trotter steps sufficient in practice. The load-bearing mechanism is parameter concentration: averaging optimized QAOA angles over random instances yields a fixed schedule that behaves like the fully optimized angles.

What would settle it

Take the fixed angles obtained from 30 random 3-regular MaxCut instances and run QAOA-for-QRAO on a different graph family, such as 4-regular graphs or random graphs with a different degree distribution at N≈30, comparing the rounded approximation ratio αc against per-instance optimized angles; if the fixed-angle αc falls noticeably below the optimized value, or if the optimized β and γ values are not tightly clustered, parameter concentration fails and the central non-variational claim collapses.

Watch

Extended reading notes

Core claim

The paper's central claim is that QRAO's space savings do not have to come with variational training. On MaxCut over random 3-regular graphs with N up to 26 and p up to 6, a fixed parameter schedule obtained by averaging optimized angles over 30 small instances performs comparably to fully optimized, per-instance parameters, in both the relaxed approximation ratio αr and the classical rounded ratio αc. A Z mixer with the |0⟩ initial state outperforms the standard X mixer and |+⟩ state for the relaxed Hamiltonian, and a Grouped Trotter implementation with T=2 or 3 Trotter steps recovers near-exact performance. The paper concludes that QRAO can be executed as a non-variational algorithm, using up to three times fewer qubits than standard QAOA while keeping good solution quality.

Load-bearing premise

The load-bearing premise is that optimal QAOA angles for the relaxed, non-diagonal QRAO Hamiltonian are tightly concentrated across random instances of a problem class, so the average over 30 small 3-regular MaxCut instances serves as a universal schedule for all instances of that class and beyond.

Editorial extensions

If this is right

  • QAOA-for-QRAO can be executed without per-instance variational training, removing the shot overhead and barren-plateau risk associated with variational parameter search.
  • The relaxed encoding cuts the qubit count by up to a factor of three for MaxCut, allowing a fixed hardware register to address larger problem instances.
  • A small number of Trotter steps (T=2 or 3) is enough for the non-diagonal cost layer, so the circuit overhead of the relaxed Hamiltonian is modest in practice.
  • The fixed parameter set transfers to larger instances than those used to choose it, and the same averaging protocol works for the (2,1)-QRAC variant.
  • Entanglement entropy grows with QAOA depth and correlates with approximation quality, suggesting that the relaxed target state uses entanglement in a way that standard QAOA on classical Hamiltonians does not.

Reading between the lines

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

  • Inference: If parameter concentration extends beyond 3-regular MaxCut, the same averaging protocol could produce fixed schedules for other NP-hard optimization families, making QRAO a drop-in space-saver for industrial solvers.
  • Inference: The observed approximation-ratio gap to standard QAOA means the threefold qubit saving is bought at some quality cost; on noisy hardware, the extra two-qubit gates from Trotterization may widen that gap, so the trade-off should be re-evaluated at realistic error rates.
  • Inference: A direct testable extension is to train fixed parameters on one graph family and benchmark on another; if concentration fails, a small per-instance fine-tuning step could recover most of the performance while keeping most of the space savings.
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

5 major / 7 minor

Summary. The manuscript proposes QAOA-for-QRAO, a fixed-parameter (non-variational) approach to Quantum Random Access Optimization, and benchmarks it for MaxCut on 3-regular graphs. The authors compare several mixers and initial states, define a parameter-setting protocol that averages optimized angles over training instances, approximate the non-diagonal cost Hamiltonian with (grouped) Trotterization, and compare the resulting approximation ratios and qubit counts with standard QAOA in noiseless and noisy simulations. The headline claims are that instance-independent fixed parameters achieve good performance without per-instance variational training and that the QRAO encoding uses up to three times fewer qubits than standard QAOA.

Significance. If the parameter-concentration behavior holds beyond the tested family, this is a useful step toward executing QRAO on early fault-tolerant devices without variational optimization overhead, with a clear resource-vs-quality trade-off. The paper is honest in acknowledging that standard QAOA gives better classical approximation ratios, and the observation that only a few Trotter steps suffice is of independent interest. However, the evidence is currently limited to random 3-regular MaxCut instances, and the exact fixed parameters are not disclosed, which limits near-term reproducibility.

major comments (5)
  1. [III-B and Fig. 4] The text in Section III-B states that the fixed parameters are obtained from 30 randomly generated instances, while Fig. 4 (left) and several other figure captions (Figs. 3, 5, 6, 8) report 120 random instances. This inconsistency must be resolved. Please clarify whether the optimization/training set contains 30 or 120 instances, and whether the evaluation in Fig. 5 is performed on the same instances used for training or on a disjoint held-out set.
  2. [III-B] The exact fixed parameter values (gamma_l, beta_l for p=1 through 6) are never tabulated. Since the central claim is that these fixed parameters perform well without per-instance optimization, the omission prevents independent verification. The authors should include a table of the fixed angles (or release code/data) so that the numerical results can be reproduced.
  3. [III-A, Fig. 3] The mixer comparison in Fig. 3 uses the parameter set "as described in Section III-B". Because those fixed parameters are obtained by averaging optimized parameters under a particular mixer (apparently the Z mixer, given the later design choice), the comparison is confounded: the X and Y mixers are evaluated with parameters that were not optimized for them. To support the claim that the Z mixer is the best choice, the authors should either optimize parameters separately for each mixer or explicitly frame Fig. 3 as a comparison within their specific fixed-parameter protocol rather than a general mixer comparison.
  4. [III-B, Fig. 4 (left)] The claim that optimized parameters are "spread in a small range" is supported only by visual inspection of a scatter plot at p=1. No quantitative measure (variance, interquartile range, or similar) is provided for p=1 through 6. Moreover, the fixed parameters are tested only on random 3-regular MaxCut graphs (up to N=26) and on the same family in the (2,1)-QRAC appendix. This validates the protocol for this specific distribution but does not substantiate the broader abstract statement that fixed parameters "remove the need for variational parameter optimization" for QRAO in general. Please qualify the claim and provide quantitative concentration statistics.
  5. [II-B and Fig. 7C] The QRAO encoding requires that adjacent vertices be placed on distinct qubits. The paper states that the number of qubits is "a third" of the standard QAOA encoding, but it does not state when this is achievable with three variables per qubit. For some 3-regular graphs, no independent set of size three exists (e.g., the triangular prism), so the idealized N/3 qubit count is not universal. The "up to three times fewer" phrasing in the conclusion is safe, but the earlier statement in Section II-B should be qualified with the conditions under which the (3,1)-QRAC encoding achieves exactly one qubit per three variables.
minor comments (7)
  1. [III-A] The sentence "the ground state of the Z mixing Hamiltonian is |0>⊗N" is inconsistent with the standard convention Z|0>=+|0>, for which the ground state of Σ_i Z_i is |1...1>. If the authors instead mean the ground state of −Σ_i Z_i, the sign convention should be stated explicitly.
  2. [III-C] The phrase "grouped two-qubit Pauli terms commute if they do not share the same Pauli string" is imprecise; such terms commute when they act on disjoint qubits (or differ by a scalar). Please rephrase for accuracy.
  3. [Algorithm 1] The notation |\hat{\psi}_p> in the Ensure line is not defined. Please state explicitly that it denotes the state produced by the Trotterized approximation of the cost evolution.
  4. [IV] The comparison in Fig. 7 uses Grouped Trotter with T=2, but the choice of T=2 is not justified. Please explain why this value was selected, especially since Fig. 6 shows a small performance gain from T=3.
  5. [Fig. 7B] The x-axis label for the noise scale appears to be missing or truncated in the figure. Please ensure that all axes are clearly labeled.
  6. [Data/Code Availability] The statements that code and data are "available upon reasonable request" are insufficient for reproducibility of the reported fixed parameters. Please provide an open-access repository or, at minimum, the exact fixed angle values.
  7. [References] Reference [9] is listed without a journal or arXiv identifier. Please complete the citation so readers can locate the source.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation: fixed-parameter claim is an empirical transfer claim with out-of-sample tests.

full rationale

The central claim is empirical rather than derivational. Fixed parameters are constructed by averaging optimized angles over a training set of 3-regular MaxCut instances (Section III-B), and the transfer claim is then tested on larger held-out instances (Fig. 5C, N up to 26) and on a different (2,1)-QRAC encoding (Appendix A). These out-of-sample tests make the claim falsifiable; the fixed schedule is not equal by construction to the per-instance optima against which it is compared. The mixer, Trotter, and noise benchmarks are comparisons run with fixed parameters, not derivations from them. Self-citations [30,34,35,38] are contextual and are not the source of the central result; the standard-QAOA fixed-angle anchor is the external result [40]. The only in-sample component is Fig. 5A/B's concentration and performance-ratio evidence on the same graph family used to calibrate the angles; this is a calibration/generalization limitation, not a definitional reduction. The text's 30-instance versus 120-instance discrepancy and the undisclosed fixed-angle values are reproducibility and correctness risks, but they do not make the derivation circular.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The central empirical claims rest on free parameters fitted during training (the fixed QAOA angles and the Trotter step), plus assumptions inherited from QRAO theory and QAOA heuristics. No invented physical entities are introduced.

free parameters (2)
  • QAOA fixed angles (gamma_l, beta_l) for l=1..p = Not reported numerically; shown in Fig. 4
    Obtained by averaging optimized parameters over 30 random 3-regular MaxCut instances (N=10..16); used for all other instances, so the central claim depends on these fitted values generalizing.
  • Trotter step T = 2
    Chosen after comparing T in {1,2,3,5,10} in Fig. 6; a hand-picked hyperparameter affecting cost-layer fidelity.
assumptions (4)
  • domain assumption The (3,1)-QRAC encoding (Eq. 8) has a decoding success probability above 1/2, and the relaxed Hamiltonian ground state encodes approximate MaxCut solutions.
    Invoked in Section II-B; the entire QRAO approach inherits this from prior work (refs. 8,9).
  • ad hoc to paper QAOA parameter concentration heuristics validated for diagonal Hamiltonians extend to non-diagonal relaxed Hamiltonians.
    Section III-B assumes this extension; the paper provides empirical evidence only for 3-regular MaxCut, no proof.
  • domain assumption First-order Trotterization with T=2 sufficiently approximates e^{-i gamma H_C} for these instances.
    Section III-C; empirically supported by Fig. 6, not derived.
  • domain assumption Adiabatic alignment intuition (initial state as ground state of the mixer) guides QAOA mixer choice.
    Section III-A uses adiabatic theorem intuition to justify mixer and initial-state choices, though QAOA dynamics differ; validated only empirically.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Non-Variational Quantum Random Access Optimization with Alternating Operator Ansatz." pith.science (2026). https://pith.science/paper/7BPHYWA4

@misc{pith2026250204277,
  author       = {Pith},
  title        = {Pith review of: Non-Variational Quantum Random Access Optimization with Alternating Operator Ansatz},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/7BPHYWA4}},
  note         = {Machine review of arXiv:2502.04277}
}
read the original abstract

Solving hard optimization problems is one of the most promising application domains for quantum computers due to the ubiquity of such problems in industry and the availability of broadly applicable quantum speedups. However, the ability of near-term quantum computers to tackle industrial-scale optimization problems is limited by their size and the overheads of quantum error correction. Quantum Random Access Optimization (QRAO) has been proposed to reduce the space requirements of quantum optimization. However, to date QRAO has only been implemented using variational algorithms, which suffer from the need to train instance-specific variational parameters, making them difficult to scale. We propose and benchmark a non-variational approach to QRAO based on the Quantum Alternating Operator Ansatz (QAOA) for the MaxCut problem. We show that instance-independent ``fixed" parameters achieve good performance, removing the need for variational parameter optimization. Additionally, we evaluate different design choices, such as various mixers, initial states, and QRAO-specific implementations of the QAOA cost operator, and identify a strategy that performs well in practice. Our results pave the way for the practical execution of QRAO on early fault-tolerant quantum computers.

Figures

Figures reproduced from arXiv: 2502.04277 by the authors.

Figure 1
Figure 1. An example of a 4-qubit QAOA circuit for problems with a classical [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 3
Figure 3. Benchmarking QAOA-for-QRAO using different mixers and initial [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figure 4
Figure 4. Visualization of fixed parameters for QAOA-for-QRAO. [PITH_FULL_IMAGE:figures/full_fig_p004_4.png] view at source ↗
Figures from the paper (5 more)
Figure 5
Figure 5. Figure 5: Performance of QAOA-for-QRAO with fixed parameters. [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]
Figure 6
Figure 6. Figure 6: Comparison of different approximated implementations of QAOA-for [PITH_FULL_IMAGE:figures/full_fig_p005_6.png]
Figure 7
Figure 7. Figure 7: Comparison between QAOA-for-QRAO and QAOA. [PITH_FULL_IMAGE:figures/full_fig_p006_7.png]
Figure 8
Figure 8. Figure 8: The evolution of entanglement entropy in QAOA-for-QRAO. We [PITH_FULL_IMAGE:figures/full_fig_p006_8.png]
Figure 9
Figure 9. Figure 9: Parameter schedule and performance of QAOA-for-QRAO with [PITH_FULL_IMAGE:figures/full_fig_p008_9.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

52 extracted references · 42 canonical work pages

  1. [1]

    A quantum algorithm for finding the minimum,

    C. Durr and P. Hoyer, “A quantum algorithm for finding the minimum,” arXiv:quant-ph/9607014, 1996

    arXiv 1996

  2. [2]

    Solving Boolean satisfiability problems with the quantum approximate optimization algorithm,

    S. Boulebnane and A. Montanaro, “Solving Boolean satisfiability problems with the quantum approximate optimization algorithm,” PRX Quantum , vol. 5, no. 3, Sep. 2024

    work page 2024

  3. [3]

    Evidence of scaling advantage for the quantum approximate optimization algorithm on a classically intractable problem,

    R. Shaydulin et al. , “Evidence of scaling advantage for the quantum approximate optimization algorithm on a classically intractable problem,” Science Advances , vol. 10, no. 22, May 2024

    work page 2024

  4. [4]

    Mind the gap: Achieving a super-grover quantum speedup by jumping to the end,

    A. M. Dalzell et al. , “Mind the gap: Achieving a super-grover quantum speedup by jumping to the end,” in Proceedings of the 55th Annual ACM Symposium on Theory of Computing . ACM, Jun. 2023

    work page 2023

  5. [5]

    Generalized short path algorithms: Towards super-quadratic speedup over Markov chain search for combinatorial optimization,

    S. Chakrabarti et al. , “Generalized short path algorithms: Towards super-quadratic speedup over Markov chain search for combinatorial optimization,” arXiv:2410.23270, 2024

    arXiv 2024

  6. [6]

    Invited: Challenges and opportunities of quantum optimization in finance,

    Z. He et al. , “Invited: Challenges and opportunities of quantum optimization in finance,” in Proceedings of the 61st ACM/IEEE Design Automation Conference , 2024

    work page 2024

  7. [7]

    Threshold for fault-tolerant quan- tum advantage with the quantum approximate optimiza- tion algorithm,

    S. Omanakuttan et al., “Threshold for fault-tolerant quan- tum advantage with the quantum approximate optimiza- tion algorithm,” arXiv preprint arXiv:2504.01897 , 2025

    arXiv 2025

  8. [8]

    Approximate solutions of combinatorial problems via quantum relaxations,

    B. Fuller et al., “Approximate solutions of combinatorial problems via quantum relaxations,” IEEE Transactions on Quantum Engineering , vol. 5, pp. 1–15, 2024

    work page 2024

Show all 52 references
  1. [9]

    Quantum-relaxation based optimiza- tion algorithms: Theoretical extensions,

    K. Teramoto et al., “Quantum-relaxation based optimiza- tion algorithms: Theoretical extensions,” 2023

    2023

  2. [10]

    Dense quantum coding and quantum finite automata,

    A. Ambainis et al. , “Dense quantum coding and quantum finite automata,” J. ACM , vol. 49, no. 4, p. 496–511, Jul. 2002

    2002

  3. [11]

    Optimal lower bounds for quantum automata and random access codes,

    A. Nayak, “Optimal lower bounds for quantum automata and random access codes,” in 40th Annual Symposium on F oundations of Computer Science (Cat. No.99CB37039) , 1999, pp. 369–376

    1999

  4. [12]

    (4,1)-quantum random access coding does not exist—one qubit is not enough to recover one of four bits,

    M. Hayashi et al., “(4,1)-quantum random access coding does not exist—one qubit is not enough to recover one of four bits,” New Journal of Physics , vol. 8, no. 8, p. 129–129, Aug. 2006

    2006

  5. [13]

    Constructions of quantum random access codes,

    T. Imamichi and R. Raymond, “Constructions of quantum random access codes,” in Asian Quantum Information Symposium , 2018

    2018

  6. [14]

    The geometry of bloch space in the context of quantum random access codes,

    L. Man ˇcinska and S. A. L. Storgaard, “The geometry of bloch space in the context of quantum random access codes,” Quantum Information Processing , vol. 21, no. 4, Apr. 2022

    2022

  7. [15]

    The role of entanglement in quantum-relaxation based optimization algorithms,

    K. Teramoto et al. , “The role of entanglement in quantum-relaxation based optimization algorithms,” in 2023 IEEE International Conference on Quantum Com- puting and Engineering (QCE) , vol. 01, 2023, pp. 543– 553

    2023

  8. [16]

    Quantum combinato- rial optimization algorithms for network reconfiguration: QRAO vs. QAOA,

    A. P. Ngo and H. T. Nguyen, “Quantum combinato- rial optimization algorithms for network reconfiguration: QRAO vs. QAOA,” in 2024 56th North American Power Symposium (NAPS), 2024, pp. 1–6

    2024

  9. [17]

    Noise robustness of quantum relax- ation for combinatorial optimization,

    K. Tamura et al. , “Noise robustness of quantum relax- ation for combinatorial optimization,” IEEE Transactions on Quantum Engineering , vol. 5, pp. 1–9, 2024

    2024

  10. [18]

    Quantum relaxation for solving multiple knapsack problems,

    M. Sharma et al. , “Quantum relaxation for solving multiple knapsack problems,” 2024

    2024

  11. [19]

    Recursive quantum relaxation for combinatorial optimization problems,

    R. Kondo et al. , “Recursive quantum relaxation for combinatorial optimization problems,” 2024

    2024

  12. [20]

    Characterizing barren plateaus in quantum ans¨atze with the adjoint representation,

    E. Fontana et al. , “Characterizing barren plateaus in quantum ans¨atze with the adjoint representation,” Nature Communications, vol. 15, no. 1, p. 7171, 2024

    2024

  13. [21]

    Diagnosing barren plateaus with tools from quantum optimal control,

    M. Larocca et al. , “Diagnosing barren plateaus with tools from quantum optimal control,” Quantum, vol. 6, p. 824, Sep. 2022

    2022

  14. [22]

    Barren plateaus in quantum neural network training landscapes,

    J. R. McClean et al., “Barren plateaus in quantum neural network training landscapes,” Nature Communications , vol. 9, no. 1, Nov. 2018

    2018

  15. [23]

    Quantum optimization,

    T. Hogg and D. Portnov, “Quantum optimization,” In- formation Sciences , vol. 128, no. 3–4, p. 181–197, Oct. 2000

    2000

  16. [24]

    A quantum approximate optimization algorithm,

    E. Farhi et al. , “A quantum approximate optimization algorithm,” arXiv preprint arXiv:1411.4028 , 2014

    2014 arXiv

  17. [25]

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

    S. Hadfield et al. , “From the quantum approximate optimization algorithm to a quantum alternating operator ansatz,” Algorithms, vol. 12, no. 2, p. 34, Feb. 2019

    2019

  18. [26]

    Efficient variational simulation of non-trivial quantum states,

    W. W. Ho and T. H. Hsieh, “Efficient variational simulation of non-trivial quantum states,” SciPost Physics, vol. 6, no. 3, Mar. 2019

    2019

  19. [27]

    Quantum alternating operator ansatz (QAOA) phase diagrams and applications for quantum chemistry,

    V . Kremenetski et al. , “Quantum alternating operator ansatz (QAOA) phase diagrams and applications for quantum chemistry,” arXiv preprint arXiv:2108.13056 , 2021

    2021 arXiv

  20. [28]

    The quantum approximate optimization algorithm at high depth for maxcut on large-girth regular graphs and the Sherrington-Kirkpatrick model,

    J. Basso et al. , “The quantum approximate optimization algorithm at high depth for maxcut on large-girth regular graphs and the Sherrington-Kirkpatrick model,” in 17th Conference on the Theory of Quantum Computation, Communication and Cryptography (TQC 2022) , vol. 232, 2022,...

    2022

  21. [29]

    Parameter setting in quantum approximate optimization of weighted problems,

    S. H. Sureshbabu et al. , “Parameter setting in quantum approximate optimization of weighted problems,” Quantum, vol. 8, p. 1231, Jan. 2024

    2024

  22. [30]

    Performance of quantum approximate opti- mization with quantum error detection,

    Z. He et al., “Performance of quantum approximate opti- mization with quantum error detection,” Communications Physics, vol. 8, no. 1, p. 217, 2025

    2025

  23. [31]

    M. A. Nielsen and I. L. Chuang, Quantum Computation and Quantum Information: 10th Anniversary Edition . Cambridge University Press, 2010

    2010

  24. [32]

    Training variational quantum algorithms is NP-hard,

    L. Bittel and M. Kliesch, “Training variational quantum algorithms is NP-hard,” Physical Review Letters , vol. 127, no. 12, Sep. 2021

    2021

  25. [33]

    The quantum approximate optimization algorithm needs to see the whole graph: A typical case,

    E. Farhi et al. , “The quantum approximate optimization algorithm needs to see the whole graph: A typical case,” 2020

    2020

  26. [34]

    Multistart methods for quantum approximate optimization,

    R. Shaydulin et al. , “Multistart methods for quantum approximate optimization,” in IEEE High Performance Extreme Computing Conference , 2019

    2019

  27. [35]

    Parameter setting heuristics make the quantum approximate optimization algorithm suitable for the early fault-tolerant era,

    Z. He et al. , “Parameter setting heuristics make the quantum approximate optimization algorithm suitable for the early fault-tolerant era,” in Proceedings of the 43rd IEEE/ACM International Conference on Computer-Aided Design, 2024, pp. 1–7

    2024

  28. [36]

    Quantum approximate optimization algorithm: Performance, mechanism, and implementation on near-term devices,

    L. Zhou et al. , “Quantum approximate optimization algorithm: Performance, mechanism, and implementation on near-term devices,” Physical Review X , vol. 10, no. 2, p. 021067, 2020

    2020

  29. [37]

    An efficient circuit compilation flow for quantum approximate optimization algorithm,

    M. Alam et al., “An efficient circuit compilation flow for quantum approximate optimization algorithm,” in 2020 57th ACM/IEEE Design Automation Conference (DAC) , 2020, pp. 1–6

    2020

  30. [38]

    Alignment between initial state and mixer improves QAOA performance for constrained optimiza- tion,

    Z. He et al. , “Alignment between initial state and mixer improves QAOA performance for constrained optimiza- tion,” npj Quantum Information , vol. 9, no. 1, p. 121, 2023

    2023

  31. [39]

    The QAOA gets stuck starting from a good classical string,

    M. Cain et al. , “The QAOA gets stuck starting from a good classical string,” arXiv preprint arXiv:2207.05089 , 2022

    2022 arXiv

  32. [40]

    Fixed-angle conjectures for the quantum approximate optimization algorithm on regular maxcut graphs,

    J. Wurtz and D. Lykov, “Fixed-angle conjectures for the quantum approximate optimization algorithm on regular maxcut graphs,” Physical Review A , vol. 104, no. 5, p. 052419, 2021

    2021

  33. [41]

    End-to-end protocol for high-quality QAOA parameters with few shots,

    T. Hao et al. , “End-to-end protocol for high-quality QAOA parameters with few shots,” arXiv preprint arXiv:2408.00557, 2024

    2024

  34. [42]

    The quantum approximate optimization algorithm at high depth for MaxCut on large-girth regular graphs and the Sherrington-Kirkpatrick model

    J. Basso et al. , “The quantum approximate optimization algorithm at high depth for MaxCut on large-girth regular graphs and the Sherrington-Kirkpatrick model.” Schloss Dagstuhl – Leibniz-Zentrum f ¨ur Informatik, 2022

    2022

  35. [43]

    O(N 3) measurement cost for varia- tional quantum eigensolver on molecular Hamiltonians,

    P. Gokhale et al. , “ O(N 3) measurement cost for varia- tional quantum eigensolver on molecular Hamiltonians,” IEEE Transactions on Quantum Engineering , vol. 1, pp. 1–24, 2020

    2020

  36. [44]

    Term grouping and travelling sales- person for digital quantum simulation,

    K. Gui et al. , “Term grouping and travelling sales- person for digital quantum simulation,” arXiv preprint arXiv:2001.05983, 2020

    2001 arXiv

  37. [45]

    Minimal universal two-qubit cnot- based circuits,

    V . V . Shende et al. , “Minimal universal two-qubit cnot- based circuits,” arXiv preprint quant-ph/0308033 , 2003

    2003 arXiv

  38. [46]

    Validating quantum computers using randomized model circuits,

    A. W. Cross et al., “Validating quantum computers using randomized model circuits,” Physical Review A, vol. 100, no. 3, p. 032328, 2019

    2019

  39. [47]

    Random compiler for fast Hamiltonian simulation,

    E. Campbell, “Random compiler for fast Hamiltonian simulation,” Physical Review Letters , vol. 123, no. 7, Aug. 2019

    2019

  40. [48]

    qujax: Simulating quantum circuits with JAX,

    S. Duffield et al. , “qujax: Simulating quantum circuits with JAX,” Journal of Open Source Software , vol. 8, no. 89, p. 5504, Sep. 2023

    2023

  41. [49]

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

    H.-Y . Huang et al. , “Predicting many properties of a quantum system from very few measurements,” Nature Physics, vol. 16, no. 10, pp. 1050–1057, 2020

    2020

  42. [50]

    How much entanglement do quantum op- timization algorithms require?

    Y . Chen et al., “How much entanglement do quantum op- timization algorithms require?” in Quantum 2.0. Optica Publishing Group, 2022, pp. QM4A–2

    2022

  43. [51]

    The quantum approximate optimiza- tion algorithm performance with low entanglement and high circuit depth,

    R. Sreedhar et al., “The quantum approximate optimiza- tion algorithm performance with low entanglement and high circuit depth,” arXiv preprint arXiv:2207.03404 , 2022

    2022 arXiv

  44. [52]

    Entanglement perspective on the quantum approximate optimization algorithm,

    M. Dupont et al. , “Entanglement perspective on the quantum approximate optimization algorithm,” Physical Review A, vol. 106, no. 2, p. 022423, 2022. DISCLAIMER This paper was prepared for informational purposes by the Global Technology Applied Research center of JPMorgan Chas...

    2022

Pith tools

Reviewed August 8, 2026 · model on record in the stance chip above.