Pith. sign in

REVIEW 4 major objections 5 minor 45 references

Transfer Learning for Deep-Unfolded Combinatorial Optimization Solver with Quantum Annealer

T0 review · 4 major / 5 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 SQA-trained step sizes transferred to a quantum annealer reach zero error within 30 iterations and cut total solving time from 34.9 s to 12.2 s.

desk verdict Classical-quantum transfer learning for the deep-unfolded Ohzeki method is a practical idea that works on the single benchmark they tried, but the evidence base is too narrow to support the conclusion's stronger claims. read the letter →

arxiv 2501.03518 v1 pith:YIWGLW6S submitted 2025-01-07 quant-ph cs.LG

Ryo Hagiwara , Shunta Arai , Satoshi Takabe This is my paper · ORCID
classification quant-phcs.LG
keywords quantumannealingdeepunfoldingOhzekimethodclassical-quantumtransferlearningcombinatorialoptimizationsimulatedimagereconstructionstep-size
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper tries to establish that a trainable quantum-annealing-based combinatorial optimization solver can be trained entirely on a classical computer and still deliver a speed advantage on real quantum hardware. It proposes classical-quantum transfer learning for the deep-unfolded Ohzeki method (DUOM), in which step sizes learned with a classical sampler are transferred to a quantum annealer. On a binary image reconstruction task, the SQA-trained version executed on a quantum annealer reached zero reconstruction error within 30 iterations and took 12.2 s total, versus 34.9 s for the fully classical SQA-trained DUOM. If correct, this removes the main obstacle to using quantum annealers inside trainable solvers: the prohibitive cost of training on quantum hardware.

What carries the argument

The load-bearing object is the deep-unfolded Ohzeki method (DUOM), a trainable version of the Ohzeki method in which the Hubbard-Stratonovich transformation converts a constrained QUBO into repeated sampling from $Q(x;v) \propto \exp(-\beta f_0(x) + \beta \sum_k v_k f_k(x))$ followed by gradient updates $v_k \leftarrow v_k + \eta_t(C_k - \langle f_k\rangle)$. Deep unfolding makes the step sizes $\eta_t$ learnable parameters, and the variance identity $\partial\langle f_k\rangle/\partial v_k = \beta(\langle f_k^2\rangle - \langle f_k\rangle^2)$ supplies gradients without differentiating through the sampler. Classical-quantum transfer learning is the scheme of training these $\eta_t$ with a classical sampler (MH or SQA) and then executing the same unfolded iterations with a quantum annealer; the Trotter number in SQA controls how closely the training sampler approximates QA.

What would settle it

Compare the empirical distribution of samples from the quantum annealer with SQA samples at the same $Q(x;v)$; if the difference in the estimated constraint expectations $\langle f_k\rangle$ is comparable to or larger than the learned step size $\eta_t$ times the gradient scale, the transferred parameters can be suboptimal or diverge. A direct version is to run SQA-QA on instances with $M/N > 0.633$ and check whether MSE still reaches zero within 30 iterations; failure would show that the transfer result depends on the easy regime.

Watch

Extended reading notes

Core claim

The central claim is that step sizes trained under simulated quantum annealing (SQA) are good enough, after transfer, for DUOM executed on a quantum annealer to solve constrained binary quadratic problems faster and more accurately than the untrained Ohzeki method with a grid-searched step size. In the experiments SQA-QA achieved the best mean squared error (MSE), reaching zero within 30 iterations, while the untrained solver did not reach zero within 50; SQA-QA also beat MH-QA, indicating that the classical sampler used for training matters. In wall-clock terms, the transferred solver took 12.2 s per instance including API latency, while the classically executed SQA-SQA took 34.9 s, even though SQA-SQA converged in fewer iterations. This is the paper's evidence that classical-quantum transfer learning can make a trainable quantum COP solver practical.

Load-bearing premise

The step sizes trained under SQA remain effective when the sampler is swapped for a real quantum annealer, which requires the hardware annealer's output distribution to be close enough to SQA's at the distributions $Q(x;v)$ encountered during execution.

Editorial extensions

If this is right

  • SQA-trained step sizes transferred to a quantum annealer reach zero MSE within 30 iterations on the tested image reconstruction instances, while the untrained Ohzeki method with constant step size does not reach zero within 50.
  • Larger Trotter numbers in SQA improve transfer: $\tau=4$ and $\tau=8$ succeed where $\tau=1$ (equivalent to MH at $\beta=1$) fails.
  • Executing DUOM on a quantum annealer reduces total wall-clock time to 12.2 s per instance versus 34.9 s for the classically executed SQA-SQA, even though SQA-SQA converges in fewer iterations.
  • Because DUOM accepts any sampler, the training phase can be kept classical, avoiding the roughly one million sampling calls per training run that quantum execution would require.

Reading between the lines

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

  • The reported speed advantage depends on the quantum annealer's sampling rate and API latency; on a different device or with local access, the wall-clock comparison could shift, and the paper's 12.2 s figure is dominated by API response time.
  • The transfer's validity rests on an untested distributional match between SQA and the hardware annealer; a systematic comparison of hardware samples with $Q(x;v)$ could select the best classical surrogate and Trotter number instead of assuming SQA is close.
  • The experiments use one fixed $15\times 15$ image with $M/N = 0.6$ below the $0.633$ threshold; testing near or above that threshold, or on inequality-constrained knapsack instances, would show whether the zero-MSE result is a property of the method or of the easy regime.
  • The same recipe—train an unfolded iterative algorithm on a classical surrogate, then deploy it on quantum hardware—should apply to other trainable quantum-classical solvers beyond DUOM, provided the surrogate's output distribution is close enough to the hardware's.
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

4 major / 5 minor

Summary. The manuscript proposes classical-quantum transfer learning for the deep-unfolded Ohzeki method (DUOM). Step sizes of the Ohzeki gradient-descent updates are trained entirely with a classical sampler -- either Metropolis-Hastings (MH) or simulated quantum annealing (SQA) -- and the trained step sizes are then transferred to DUOM executed on a D-Wave quantum annealer. The method is tested on a 15x15 binary image reconstruction problem with N=225 linear measurements and M=135 constraints. Numerical results show that SQA-trained DUOM executed on QA (SQA-QA) reaches zero MSE within 30 iterations, outperforms MH-trained transfer and a fixed-step-size grid-search baseline, and has lower total execution time than SQA-SQA. The paper also studies the effect of the SQA Trotter number and concludes that SQA is a more suitable classical surrogate than MH.

Significance. If the transfer premise is valid, the proposal is practically valuable: it removes quantum computation from the training loop while retaining quantum hardware at inference, and it extends deep unfolding to quantum samplers. The paper is commendably concrete: it reports error bars, compares multiple samplers, includes a grid-search baseline, and evaluates execution time, not just iterations. The SQA-vs-MH comparison is a sensible way to probe surrogate quality, and the claim that SQA-trained parameters transfer better than MH-trained parameters is an informative empirical result. However, the central validation rests on a single fixed ground-truth image and on an uncalibrated assumption that D-Wave sampling statistics are close enough to the SQA target distribution; both need to be addressed before the broader conclusions can be accepted.

major comments (4)
  1. [Section IV and Section V.B] The transfer premise is asserted but never directly calibrated. Section IV states that parameters trained with a classical sampler can be used with a quantum annealer if the difference is sufficiently small, but the manuscript provides no measurement of the difference between D-Wave samples and the theoretical distribution Q(x; v(t)) used in the update rule (4) and the stability factor (7). If <f_k>_QA is biased relative to <f_k>_Q, the step sizes trained under SQA may be suboptimal or cause divergence. The paper should either compare empirical constraint expectations <f_k> obtained from D-Wave and SQA at the trained v(t) trajectory, or report a sensitivity analysis over D-Wave sampling parameters (anneal time, number of reads, chain strength, postprocessing), none of which are currently reported.
  2. [Section V.A and Figures 4-6] All experiments use one fixed ground-truth image x* (Fig. 3), with the dataset consisting of random matrices A and corresponding observations y = A x*. Thus the headline 'MSE reaches zero within 30 iterations' is an average over 50 instances of a single image. This limits the support for the general claim that the learned step sizes accelerate convergence and for the conclusion's statement of 'practical viability'. The authors should test on multiple independent ground-truth images or otherwise justify that the learned step sizes are not image-specific.
  3. [Section V.B.2 and Figure 5] The Trotter-number experiment is confounded by the simultaneous scaling of the inverse temperature: the figure reports results for (beta, tau) = (1,1), (4,4), and (8,8). Because beta itself controls the sharpness of Q(x;v), the improved performance with larger tau cannot be attributed to a better approximation of QA dynamics; it may simply reflect a lower-temperature target distribution. The authors should fix beta (or vary it systematically) while changing the Trotter number to isolate the effect of the SQA approximation.
  4. [Section V.B.3 and Table I] The execution-time comparison is not fully reproducible as reported. The manuscript does not state the number of D-Wave API calls (presumably one per iteration, so 30 calls for SQA-QA), the per-call anneal time, the number of reads, or the embedding parameters that produce the reported 181 ms sampling time. The 12.2 s API column dominates the SQA-QA total, so the '190 times faster' statement applies only when API latency is ignored, while the practical speed-up is 2.9x. The authors should report the API call structure and hardware settings so that the wall-clock comparison can be reproduced and interpreted fairly.
minor comments (5)
  1. [Section IV] There is a duplicated phrase 'without without QA during the training process'; please revise.
  2. [Section III.B] The text contains several typos ('data-riven', 'e fficient', 'Ohezeki') that should be corrected in a final pass.
  3. [Equation (6) and (7)] The loss function is denoted L_lambda in Eq. (2) and the text, but Eq. (6) uses L without a subscript; please make the notation consistent.
  4. [Section V.B.1] The sentence 'SQA-QA achieved the highest performance' should be qualified, since Figure 6 shows that SQA-SQA converges faster than SQA-QA; the claim is true only among the QA-executed methods compared in Figure 4.
  5. [General] No code or data availability statement is included. Given that the paper is empirical and relies on specific random instances and hardware settings, providing code and the exact instance generation parameters would substantially improve reproducibility.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the transfer result is an empirical experiment, not a re-statement of fitted inputs or a self-citation-forced conclusion.

full rationale

The derivation chain is not circular. The only load-bearing dependencies from the authors' prior work are the DUOM construction and the variance-based gradient estimator (Eqs. (6)-(8)), which are cited to Ref. [13] and are externally testable published algorithms; the paper does not invoke them as a uniqueness theorem or to forbid alternatives. The central new claim, classical-quantum transfer learning, is an experimental protocol: step sizes are trained classically (MH or SQA) and then executed on a D-Wave annealer, with performance compared against grid-searched Ohzeki, MH-MH, and SQA-SQA baselines on random instances. The reported MSE is the reconstruction error of sampler outputs, not a re-statement of the training loss (Eq. (2)); the execution-time advantage is a measured hardware result. The transfer premise that the SQA-QA difference is 'sufficiently small' is explicitly framed as an assumption and tested indirectly through Trotter-number scaling (Fig. 5); an unvalidated empirical assumption is a correctness risk, not a circular step. No equation in the paper is defined in terms of the quantity it is used to predict, and no fitted parameter is renamed as a prediction. Self-citation to Ref. [13] is present, but it supplies the starting algorithm rather than forcing the new transfer result, so no circular step meets the evidentiary bar of Eq. X = Eq. Y by construction.

Assumptions & free parameters 6 free parameters · 5 assumptions · 0 invented entities

The central claim rests on standard statistical mechanics identities and on two unverified similarities: between the quantum annealer and the classical Boltzmann distribution assumed by the method, and between SQA and the actual QA hardware. All numerical hyperparameters are chosen by hand, and the paper provides no sensitivity analysis.

free parameters (6)
  • Inverse temperature beta = not reported for D-Wave runs; set beta=tau in SQA experiments
    The sampling distribution Q(x;v) depends on beta; the paper does not state what beta was used with the quantum annealer.
  • Penalty strength lambda = 1
    Used in the loss function (Eq. 2) for training and evaluation; no sensitivity analysis.
  • Initial learning rate (Adam) = 5.0e-2, decayed by factor 0.8 each iteration
    Training hyperparameter for step-size optimization; chosen by hand.
  • Number of unfolded iterations T = 30
    Total iterations in DUOM; results for baselines extend to 50.
  • Mini-batch configuration = 20 mini-batches of size 4
    Training procedure detail; chosen by hand.
  • Trotter number tau = 4 (default); tested 1, 4, 8
    Controls SQA accuracy; larger tau improves transfer performance according to Section V.B.2.
assumptions (5)
  • standard math Hubbard-Stratonovich transformation linearizes the quadratic penalty, yielding the sampling distribution Eq. (3).
    The Ohzeki method relies on this transformation (citations [10-12]) to avoid additional qubits; the derivation is standard.
  • standard math The derivative of the expectation with respect to v_k equals beta times the variance, Eq. (8).
    Used for gradient estimation; a standard statistical identity for Boltzmann distributions.
  • domain assumption The image reconstruction problem with noiseless linear measurements and a fixed original image x* is representative of constrained COPs.
    All experiments use one fixed ground-truth image of size 15x15 with random measurement matrices A; generalization to other images or problem types is not tested.
  • domain assumption D-Wave's samples approximate the classical Boltzmann distribution Q(x;v) closely enough for the gradient update (Eq. 4) to be valid.
    The Ohzeki method's convergence relies on accurate expectation estimates. The paper does not calibrate the annealer's effective temperature or sampling bias.
  • domain assumption The SQA sampler approximates the D-Wave annealer's distribution closely enough for transfer learning to preserve step-size quality.
    This is the core transfer-learning premise, tested only indirectly via Trotter number in Section V.B.2.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Transfer Learning for Deep-Unfolded Combinatorial Optimization Solver with Quantum Annealer." pith.science (2026). https://pith.science/paper/YIWGLW6S

@misc{pith2026250103518,
  author       = {Pith},
  title        = {Pith review of: Transfer Learning for Deep-Unfolded Combinatorial Optimization Solver with Quantum Annealer},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/YIWGLW6S}},
  note         = {Machine review of arXiv:2501.03518}
}
read the original abstract

Quantum annealing (QA) has attracted research interest as a sampler and combinatorial optimization problem (COP) solver. A recently proposed sampling-based solver for QA significantly reduces the required number of qubits, being capable of large COPs. In relation to this, a trainable sampling-based COP solver has been proposed that optimizes its internal parameters from a dataset by using a deep learning technique called deep unfolding. Although learning the internal parameters accelerates the convergence speed, the sampler in the trainable solver is restricted to using a classical sampler owing to the training cost. In this study, to utilize QA in the trainable solver, we propose classical-quantum transfer learning, where parameters are trained classically, and the trained parameters are used in the solver with QA. The results of numerical experiments demonstrate that the trainable quantum COP solver using classical-quantum transfer learning improves convergence speed and execution time over the original solver.

Figures

Figures reproduced from arXiv: 2501.03518 by the authors.

Figure 1
Figure 1. FIG. 1: A schematic diagram of deep-unfolded gradient [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2: Overview of classical-quantum transfer learning, [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. FIG. 3: Examples of reconstructed images using the Ohzeki [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: FIG. 4: MSE performance of DUOM by classical-quantum [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: FIG. 5: Trotter number [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

45 extracted references · 39 canonical work pages

  1. [1]

    Performance of DUOM with quantum annealer First, we evaluated the performance of classical-quantum transfer learning. We compared three methods that uses QA in their execution: DUOM trained using SQA (SQA-QA), DUOM trained using MH (MH-QA), and the original Ohzeki method with a constant step size was optimized by a grid search. Figure 4 illustrates perfor...

    work page

  2. [2]

    The Trotter number τ is a hyperparameter that controls the accuracy of the approximation

    Trotter number dependency in SQA SQA approximates QA by using multiple Trotter layers, as introduced by the Suzuki-Trotter decomposition [40, 41]. The Trotter number τ is a hyperparameter that controls the accuracy of the approximation. Given a limit of τ → ∞, SQA reproduces the original quantum system described by the Schr ¨odinger equation, as required ...

    work page

  3. [3]

    5: Trotter numberτ dependency of MSE performance for SQA-QA whenβ =τ

    Comparison of quantum DUOM with classical DUOM We now compare DUOM performance with different sam- plers used during execution We considered three scenarios: 6 0 10 20 30 40 50 Iteration 10 4 10 3 10 2 10 1 MSE = 1, = 1 = 4, = 4 = 8, = 8 FIG. 5: Trotter numberτ dependency of MSE performance for SQA-QA whenβ =τ. DUOM trained using SQA and executed on QA (S...

    work page

  4. [4]

    The Complexity of Theorem-Proving Proce- dures,

    S. A. Cook, “The Complexity of Theorem-Proving Proce- dures,” in Proceedings of the Third Annual ACM Symposium on Theory of Computing (ACM, New York, 1971), pp. 151–158

    work page 1971

  5. [5]

    Kirkpatrick, C

    S. Kirkpatrick, C. D. Gelatt, and M. P. Vecchi, Science220, 671 (1983)

    1983

  6. [6]

    Kadowaki and H

    T. Kadowaki and H. Nishimori, Phys. Rev. E 58, 5355 (1998)

    1998

  7. [7]

    V . S. Denchev, S. Boixo, S. V . Isakov, N. Ding, R. Babbush, V . Smelyanskiy, J. Martinis and H. Neven, Phys. Rev. X 6, 031015 (2016)

    work page 2016

  8. [8]

    D-Wave Systems Inc., https://www.dwavesys.com

    work page

Show all 45 references
  1. [9]

    Choi, Quant

    V . Choi, Quant. Inf. Proc.7, 193 (2008)

    2008

  2. [10]

    Choi, Quant

    V . Choi, Quant. Inf. Proc.10, 343 (2001)

    2001

  3. [11]

    Klymko, B

    C. Klymko, B. D. Sullivan, and T. S. Humble, Quantum Infor- mation Processing 13, 709 (2014)

    2014

  4. [12]

    Vinci, T

    W. Vinci, T. Albash, G. Paz-Silva, I. Hen, and D. A. Lidar, Phys. Rev. A 92, 042310 (2015)

    2015

  5. [13]

    Ohzeki, Sci

    M. Ohzeki, Sci. Rep. 10, 3126 (2020)

    2020

  6. [14]

    R. L. Stratonovich, Sov. Phys. Dokl. 2, 416 (1957)

    1957

  7. [15]

    Hubbard, Phys

    J. Hubbard, Phys. Rev. Lett. 3, 77 (1959)

    1959

  8. [16]

    Hagiwara and S

    R. Hagiwara and S. Takabe, J. Phys. Soc. Jpn. 93, 063801 (2024)

    2024

  9. [17]

    Learning Fast Approximations of Sparse Coding,

    K. Gregor and Y . LeCun, “Learning Fast Approximations of Sparse Coding,” in Proceedings of the 27th International Con- ference on Machine Learning (Omnipress, 2010), pp. 399–406

    2010

  10. [18]

    J. R. Hershey, J. Le Roux, and F. Weninger: Deep Unfold- ing: Model-Based Inspiration of Novel Deep Architectures; arXiv:1409.2574

    arXiv

  11. [19]

    Takabe and T

    S. Takabe and T. Wadayama, IEICE Transactions on Funda- mentals of Electronics, Communications and Computer Sci- ences, 105, 1110 (2022)

    2022

  12. [20]

    Deep Unfolding for Communications Systems: A Survey and Some New Direc- tions,

    A. Balatsoukas-Stimming and C. Studer, “Deep Unfolding for Communications Systems: A Survey and Some New Direc- tions,” in Proceedings of the 2019 IEEE International Work- shop on Signal Processing Systems (IEEE, 2019), pp. 266–271

    2019

  13. [21]

    D. Ito, S. Takabe, and T. Wadayama, IEEE Trans. Signal Pro- cess. 67, 3113 (2019)

    2019

  14. [22]

    Deep Unfolding Network for Image Super-Resolution,

    K. Zhang, L. Van Gool, and R. Timofte, “Deep Unfolding Network for Image Super-Resolution,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (IEEE, 2020), pp. 3217–3226

    2020

  15. [23]

    Takabe and T

    S. Takabe and T. Abe, IEEE Wireless Commun. Lett. 13, 701 (2024)

    2024

  16. [24]

    Takabe, arXiv:2306.16264 (2023)

    S. Takabe, arXiv:2306.16264 (2023)

    2023 arXiv

  17. [25]

    Arai and S

    S. Arai and S. Takabe, arXiv:2408.03026, accepted to Phys. Rev. Res. (2024)

    2024 arXiv

  18. [26]

    Metropolis, A

    N. Metropolis, A. W. Rosenbluth, M. N. Rosenbluth, A. H. Teller and E. Teller, J. Chem. Phys.,21, pp. 1087–1092 (1953)

    1953

  19. [27]

    W. K. Hastings, Biometrika 57, 97 (1970)

    1970

  20. [28]

    Simulated Quantum Annealing Can Be Exponentially Faster than Classical Simulated Anneal- ing,

    E. Crosson and A. W. Harrow, “Simulated Quantum Annealing Can Be Exponentially Faster than Classical Simulated Anneal- ing,” in Proceedings of the 57th Annual IEEE Symposium on F oundations of Computer Science(IEEE, 2016), pp. 714–723

    2016

  21. [29]

    G. E. Santoro, R. Marton´ak, E. Tosatti, and R. Car, Science295, 2427 (2002)

    2002

  22. [30]

    Glover, G

    F. Glover, G. Kochenberger, R. Hennig, and Y . Du, Ann. Oper. Res. 314, 141 (2022)

    2022

  23. [31]

    Monga, Y

    V . Monga, Y . Li, and Y . C. Eldar, IEEE Signal Process. Mag. 38, 18 (2021)

    2021

  24. [32]

    Shlezinger, Y

    N. Shlezinger, Y . C. Eldar, and S. P. Boyd, IEEE Access 10, 115384 (2022)

    2022

  25. [33]

    Takabe and T

    S. Takabe and T. Wadayama, IEICE Trans. Fundam. Electron. Commun. Comput. Sci. E105-A, 1110 (2022)

    2022

  26. [34]

    D. P. Kingma and J. Ba, arXiv:1412.6980 (2017)

    2017 arXiv

  27. [35]

    PyTorch, https://pytorch.org

  28. [36]

    Blekos, D

    K. Blekos, D. Brand, A. Ceschini, C. H. Chou, R. H. Li, K. Pandya, and A. Summer, Phys. Rep. 1068, 1 (2024)

    2024

  29. [37]

    Wierichs, J

    D. Wierichs, J. Izaac, C. Wang, and C. Lin, Quantum 6, 677 (2022)

    2022

  30. [38]

    Dborin, F

    J. Dborin, F. Barratt, V . Wimalaweera, L. Wright, and A. G. Green, Quantum Sci. Technol. 7, 035014 (2022)

    2022

  31. [39]

    Evenbly and G

    G. Evenbly and G. Vidal, J. Stat. Phys. 145, 891 (2011)

    2011

  32. [40]

    Tanaka, IEEE Trans

    T. Tanaka, IEEE Trans. Inf. Theory 48, 2888 (2002)

    2002

  33. [41]

    Takabe, M

    S. Takabe, M. Imanishi, T. Wadayama, and K. Hayashi, IEEE Access 7, 93326 (2019)

    2019

  34. [42]

    OpenJij, https: //www.openjij.org/

  35. [43]

    H. F. Trotter, Proc. Am. Math. Soc. 10, 545 (1959). 8

    1959

  36. [44]

    Suzuki, Commun

    M. Suzuki, Commun. Math. Phys. 51, 183 (1976). [42] Y . Bando and H. Nishimori, Phys. Rev. A 104, 022607 (2021)

    1976

  37. [45]

    Yu and T

    S. Yu and T. Nabil, Front. Phys. 9, 659463 (2021)

    2021

Pith tools

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