Pith. sign in

REVIEW 3 major objections 5 minor 1 cited by

A quantum annealer, used as ground truth, shows a tensor-network method's error stays flat where theory predicted a drop.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

Using QA hardware as a proxy ground truth, the authors find BP-TNS correlation error does not decrease from L=3 to L=4 cubic dimer lattices, contrary to the prediction in Ref [10].

T0 review reviewed 2026-08-05 challenge →

load-bearing objection A useful, honest reverse-benchmarking paper that probably kills a specific BP-TNS scaling prediction, but the key L=4 estimate leans on an untested QA-error-size-constancy assumption. the 3 major comments →

arxiv 2508.15759 v1 pith:RQVEMR65 submitted 2025-08-21 quant-ph

Evaluating classical simulations with a quantum processor

Alberto Nocera , Jack Raymond , William Bernoudy , Mohammad H. Amin , Andrew D. King This is my paper
classification quant-ph
keywords quantum annealingtensor networksbelief propagationspin glassquench dynamicsquantum simulation benchmarkingmatrix product states
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

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 asks whether a quantum annealing processor can serve as a trusted referee for classical simulations of quantum dynamics when classical ground truth becomes too expensive to compute. It uses a quantum processor to estimate the two-body correlation error of belief-propagation tensor-network (BP-TNS) simulations of Ising spin-glass quenches on cubic dimer lattices up to 128 qubits. The central finding is that BP-TNS error does not decrease from L=3 to L=4 despite the disappearance of three-site periodic loops, contradicting a recent scaling prediction. If correct, this warns against assuming that asymptotic scaling of classical simulation accuracy sets in at accessible system sizes, and it demonstrates a role for quantum processors in benchmarking classical algorithms rather than the reverse.

Core claim

The authors generate an approximate ground truth with a quantum annealing processor and use it to estimate the BP-TNS error relative to a matrix-product-state ground truth via a quadrature subtraction: the squared BP-TNS-versus-QA error minus the squared QA-versus-MPS error, under assumptions of statistical independence between the two error sources and constancy of QA error with system size. Applying this estimator at L=4 (N=128), beyond the reach of practical MPS ground truths, they find no evidence that the median BP-TNS correlation error decreases from L=3 to L=4 over the ensembles tested. For bimodal spin glasses the estimated error continues to rise; for uniform spin glasses it is flat

What carries the argument

The central object is the estimator of Eq. (5), which computes the BP-TNS error relative to an MPS ground truth as the quadrature difference between BP-TNS error relative to a quantum annealer and the annealer's own error relative to MPS. It relies on two assumptions: that QA and BP-TNS two-body correlation errors are statistically independent, and that QA error is roughly constant from L=3 to L=4. This estimator is what allows the paper to push BP-TNS error measurements to system sizes where MPS ground truth would require bond dimension around 3000.

Load-bearing premise

The load-bearing premise is that the quantum processor's own error stays the same from L=3 to L=4, but that constancy was only demonstrated on square lattices, not on the cubic dimer lattices with bimodal couplings used here; if QA error grows with system size, the subtraction understates the tensor-network error and could mask the predicted drop.

What would settle it

Run the same quench on a handful of L=4 cubic dimer instances and compute an MPS ground truth at large enough bond dimension for convergence; if the directly measured BP-TNS error decreases from L=3 to L=4, or if QA error measured this way rises, the paper's conclusion fails.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

If this is right

  • Loop corrections in BP-TNS should not be assumed to generically improve estimates; they increase error for bimodal couplings and longer quenches.
  • Increasing bond dimension in BP-TNS offers diminishing returns in loop-rich lattices, and measurement time appears exponential over the range studied rather than the asymptotic χ^7.
  • A quantum processor can act as a ground-truth source for benchmarking classical algorithms in regimes where classical ground truth is impractical.
  • The predicted asymptotic regime for BP-TNS error does not govern at N=128, so scaling predictions for these algorithms need to be treated with caution or tested at even larger sizes.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • If QA error actually grows with system size in cubic dimer lattices, the quadrature subtraction would understate BP-TNS error at L=4 and could mask a genuine decrease; this is testable with high-bond-dimension MPS on a small set of L=4 instances.
  • The dimer-perfect-correlation approximation used to expand correlation measurements is valid on average (mean 0.991) but has outliers down to 0.7 in bimodal inputs, which could bias bimodal error estimates.
  • The same benchmarking protocol could be applied to other approximate classical methods, such as PEPS variants or neural quantum states, wherever their errors are statistically independent of the processor's.
  • The flat uniform-ensemble behavior at ta=20 ns may indicate that the fixed-correlation-length asymptotic regime sets in only at larger N, or that loop removal is offset by other sources of error.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper proposes to use a quantum annealing (QA) processor as an approximate ground truth for evaluating the accuracy of belief-propagation-gauged tensor-network state (BP-TNS) simulations of quench dynamics in three-dimensional disordered Ising models. Since an MPS ground truth is impractical at the target system size (L=4, N=128), the authors introduce an indirect estimator, Eq. (5), that subtracts the QA error in quadrature from the BP-TNS-vs-QA error, relying on statistical independence of the two error sources and constancy of the QA error with system size. The estimator is validated at L=3 (Fig. 3). Applying it at L=4, the authors find that the estimated BP-TNS error does not decrease from L=3 to L=4, and for bimodal inputs it increases, contradicting the asymptotic-scaling prediction of Ref. [10].

Significance. If the result holds, it is a valuable cautionary result for the tensor-network community: asymptotic scaling predictions for BP-TNS in three dimensions may not govern at N=128, and QA hardware can serve as a benchmarking ground truth in regimes where classical ground truth is inaccessible. The paper's strengths include the empirical validation of the indirect estimator at L=3, the use of bootstrap confidence intervals, and the explicit statement of assumptions. The central claim, however, rests on assumptions that are only partially tested, most importantly the size-independence of the QA error on the cubic dimer geometry.

major comments (3)
  1. [Section III, Eq. (5), Fig. 4] The L=4 values of the indirect estimator are computed by setting \hat\epsilon^QA_MPS equal to the L=3 median value of \epsilon^QA_MPS. The constancy of QA error with system size is cited only to square-lattice data in Ref. [4] (Fig. S25), not to cubic dimer lattices or bimodal couplings. If the QA error grows with system size on this geometry, the quadrature subtraction in Eq. (5) inflates \hat\epsilon^TNS_MPS at L=4 and could mask the predicted decrease. Please provide a sensitivity analysis: report the critical value of \epsilon^QA_MPS(L=4) at which the observed flat or increasing trend would become a decrease, and present the available \epsilon^QA_MPS versus L data for the cubic dimer geometry. Without this, the central disconfirmation claim is not fully supported.
  2. [Section III, Fig. 4] The conclusion 'There is no evidence of \hat\epsilon^TNS_MPS decreasing from L=3 to L=4 ... This disagrees with the predictions of Ref. [10]' treats a null result as a disconfirmation. A null result only contradicts a prediction if the confidence interval excludes the predicted decrease. The predicted magnitude is not quoted, and no confidence interval for the L=3-to-L=4 difference is given. Please quantify the change (e.g., difference with CI) and compare it explicitly with the Ref. [10] prediction, or soften the claim to 'no evidence' without asserting disagreement.
  3. [Supplementary VII] The dimer-perfect-correlation approximation is applied to bimodal inputs, where the authors note ground-truth dimer correlations as low as 0.7; the mean value 0.991 is quoted only for L=3 at ta=20ns. Since the central claim compares L=3 and L=4 across ta up to 40ns, a size- or time-dependent bias in this approximation could propagate into \hat\epsilon^TNS_MPS. Please report mean and minimum dimer correlations for all ensembles used (including L=4 if available), or compute exact correlation sets for a subset of instances to bound the bias.
minor comments (5)
  1. [Supplementary VII] Typo: 'significantly effect estimates' should be 'significantly affect estimates'.
  2. [Fig. 2 caption] Typo: 'Ast a increases' should be 'As t_a increases'.
  3. [Section III] The notation \epsilon^TNS_MPS and \hat\epsilon^TNS_MPS is easy to confuse; consider renaming the indirect estimator (e.g., \tilde\epsilon^TNS_MPS) or adding a short glossary.
  4. [Section II] The term 'biclique' is used without definition; please define it for readers not familiar with D-Wave embedding terminology.
  5. [Eq. (3)] State explicitly that the sums in the normalized l2 norm run over all pairs i<j and that the denominator uses ground-truth correlations; this is clear but could be stated for precision.

Circularity Check

0 steps flagged

No circular reduction: the L=3-to-L=4 comparison is measured; the QA-error constancy assumption is a stated limitation, not a built-in conclusion.

full rationale

The central claim is a measured scaling comparison, not a fitted or definitional identity. Equation (5) is an estimator for BP-TNS error against MPS, combining measured epsilon_TNS_QA with an assumed hat-epsilon_QA_MPS taken from the largest available MPS-validated size (L=3). The estimator is validated at L=3 in Fig. 3 against direct MPS ground truth. The L=4 value is then computed from newly measured epsilon_TNS_QA data, so the observed flatness/increase in Fig. 4 is not forced by the estimator; it depends on the actual measured values. The paper explicitly flags the key assumption: 'The second assumption is that epsilon_QA_MPS is approximately constant as a function of system size in QA,' and cites Ref. [4] for square-lattice evidence. This is an extrapolation to cubic dimer/bimodal ensembles and a genuine robustness limitation, but it is not a circular equivalence: a larger QA error at L=4 would inflate the estimate, yet that is a bias scenario, not a logical identity. Self-citations to Ref. [4] are independent published evidence with MPS-ground-truth validation, so they do not constitute load-bearing circularity. The dimer-perfect-correlation approximation in Supplementary VII is acknowledged and is a modeling simplification, not a redefinition of the target. No step in the derivation reduces to its own inputs: the prediction attributed to Ref. [10] is refuted by measurement, and Eq. (5) is not used to define the observed trend into existence. Score 2 reflects the presence of minor self-citation and the unverified size-independence assumption, which are correctness risks, not circular steps.

Axiom & Free-Parameter Ledger

2 free parameters · 5 axioms · 0 invented entities

No new physical entities are introduced. The free parameters are the hand-tuned calibration scale α and the fixed BP-TNS bond dimensions. The load-bearing domain assumptions are the statistical independence of QA and BP-TNS errors, the size-constancy of QA error, and the dimer-perfect-correlation approximation, each only partially validated outside the regime where they are used.

free parameters (2)
  • energy scale α = 0.4 to 1.0 (chosen by hand depending on ta and coupling distribution)
    Supplementary VI: α multiplies the spin-glass Hamiltonian to map physical QPU anneal time to model time ta; the choice affects QA error level and therefore the QA ground truth accuracy used in the estimator.
  • BP-TNS bond dimension χ = 8 (cubic dimer), 16 (diamond)
    Section III and Fig 2: chosen to match the settings of Ref [10]; fixed across sizes, so it is a choice rather than a fitted value, but it determines the BP-TNS error level being tested.
axioms (5)
  • domain assumption QA two-body correlation errors are statistically independent of BP-TNS errors (correlation coefficient 0.07 for uniform at ta=20ns)
    Section III, Eq (4): needed for quadrature addition of squared errors; tested only for one ensemble and time point.
  • domain assumption QA error ϵ_QA_MPS is approximately constant with system size
    Section III, used to replace the unmeasured L=4 QA error by the L=3 median in Eq (5); support cited from Ref [4] Fig S25 for square lattices up to 90 sites, not cubic dimer or bimodal inputs.
  • standard math MPS-TDVP with high bond dimension provides a valid ground truth at accessible sizes
    Section II: the basis for benchmarking both QA and BP-TNS below L=4; controlled approximation error is the stated rationale.
  • domain assumption Trotterized evolution to s=0.6 is equivalent to full quench s=1.0 within error
    Supplementary VII: cited to Ref [4] SM Fig S23; used to reduce computational cost of BP-TNS and MPS runs.
  • ad hoc to paper Two spins in a dimer are perfectly correlated when expanding the measured correlation set
    Supplementary VII: reduces cost by factor 4; authors acknowledge it fails for some bimodal inputs where true dimer correlations reach 0.7, with validation only at L=3, 20ns.

reviewed 2026-08-05 · how reviews work

0 comments
Cite this review

Pith. "Pith review of Evaluating classical simulations with a quantum processor." pith.science (2026). https://pith.science/paper/RQVEMR65

@misc{pith2026250815759,
  author       = {Pith},
  title        = {Pith review of: Evaluating classical simulations with a quantum processor},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RQVEMR65}},
  note         = {Machine review of arXiv:2508.15759}
}
Share X Bluesky LinkedIn Reddit HN
read the original abstract

As simulations of quantum systems cross the limits of classical computability, both quantum and classical approaches become hard to verify. Scaling predictions are therefore based on local structure and asymptotic assumptions, typically with classical methods being used to evaluate quantum simulators where possible. Here, in contrast, we use a quantum annealing processor to produce a ground truth for evaluating classical tensor-network methods whose scaling has not yet been firmly established. Our observations run contrary to previous scaling predictions, demonstrating the need for caution when extrapolating the accuracy of classical simulations of quantum dynamics. Our results demonstrate that the virtuous cycle of competition between classical and quantum simulations can lend insight in both directions.

Figures

Figures reproduced from arXiv: 2508.15759 by Alberto Nocera, Andrew D. King, Jack Raymond, Mohammad H. Amin, William Bernoudy.

Figure 1
Figure 1. Figure 1: FIG. 1 [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2 [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: FIG. 3 [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: shows scaling of different errors with size, includ￾ing the QPU-estimated ensemble-median error ˆϵ TNS MPS. As expected, ˆϵ TNS MPS agrees with ϵ TNS MPS in most points, espe￾cially when quantum and classical errors are not close. In all cases, ˆϵ TNS MPS increases with N, except for the uni￾form case at ta = 20 ns for which the last two points are flat. There is no evidence of ˆϵ TNS MPS decreasing from L… view at source ↗
Figure 5
Figure 5. Figure 5: FIG. 5 [PITH_FULL_IMAGE:figures/full_fig_p006_5.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Comment on "Beyond-classical computation in quantum simulation"

    quant-ph 2026-07 conditional novelty 3.0

    With Monte Carlo noise and autocorrelation accounted for, Neural Quantum States achieve lower correlation error than the QPU on the studied 2D spin-glass annealing instances.

Reference graph

Works this paper leans on

39 extracted references · 23 canonical work pages · cited by 1 Pith paper · 1 internal anchor

  1. [1]

    Preskill, Quantum computing and the entanglement frontier (2012), arXiv:1203.5813

    J. Preskill, Quantum computing and the entanglement frontier (2012), arXiv:1203.5813

    Pith/arXiv arXiv 2012

  2. [2]

    A. J. Daley, I. Bloch, C. Kokail, S. Flannigan, N. Pearson, M. Troyer, and P. Zoller, Practical quantum advantage in quantum simulation, Nature607, 667 (2022)

    2022

  3. [3]

    Morvan, B

    A. Morvan, B. Villalonga, X. Mi, S. Mandr` a, A. Bengts- son,et al., Phase transitions in random circuit sampling, Nature634, 328 (2024)

    work page 2024

  4. [4]

    A. D. King, A. Nocera, M. M. Rams, J. Dziarmaga, R. Wiersema,et al., Beyond-classical computation in quantum simulation, Science388, 199 (2025)

    work page 2025

  5. [5]

    T. I. Andersen, N. Astrakhantsev, A. H. Karamlou, J. Berndtsson, J. Motruk,et al., Thermalization and crit- icality on an analogue–digital quantum simulator, Nature 638, 79 (2025)

    work page 2025

  6. [6]

    Haghshenas, E

    R. Haghshenas, E. Chertkov, M. Mills, W. Kadow, S.-H. Lin,et al., Digital quantum magnetism at the frontier of classical simulations (2025), arXiv:2503.20870

    Pith/arXiv arXiv 2025

  7. [7]

    Berezutskii, M

    A. Berezutskii, M. Liu, A. Acharya, R. Ellerbrock, J. Gray,et al., Tensor networks for quantum comput- ing, Nature Reviews Physics 10.1038/s42254-025-00853-1 (2025)

    doi:10.1038/s42254-025-00853-1 2025

  8. [8]

    Tindall and D

    J. Tindall and D. Sels, Confinement in the Transverse Field Ising Model on the Heavy Hex Lattice, Physical 6 5 10 15 10−2 10−1 100 5 10 15 10−2 10−1 100 5 10 15 10−2 10−1 100 5 10 15 10−2 10−1 100 5 10 15 10−2 10−1 100 5 10 15 10−2 10−1 100 5 10 15 101 102 103 104 lmax = 7 lmax = 4 lmax = 0 χχχ χ εc εc Seconds per two-body measurement (12 cores) εQA MPS ε...

    work page 2024

  9. [9]

    Beguˇ si´ c, J

    T. Beguˇ si´ c, J. Gray, and G. K.-L. Chan, Fast and con- verged classical simulations of evidence for the utility of quantum computing before fault tolerance, Science Ad- vances10, eadk4321 (2024)

    2024

  10. [10]

    Tindall, A

    J. Tindall, A. Mello, M. Fishman, M. Stoudenmire, and D. Sels, Dynamics of disordered quantum systems with two- and three-dimensional tensor networks (2025)

    work page 2025

  11. [11]

    G. Park, J. Gray, and G. K.-L. Chan, Simulating quan- tum dynamics in two-dimensional lattices with tensor network influence functional belief propagation (2025), arXiv:2504.07344 [quant-ph]

    arXiv 2025

  12. [12]

    M. S. Rudolph and J. Tindall, Simulating and Sampling from Quantum Circuits with 2D Tensor Networks (2025), arXiv:2507.11424

    arXiv 2025

  13. [13]

    Boixo, S

    S. Boixo, S. V. Isakov, V. N. Smelyanskiy, R. Babbush, N. Ding, Z. Jiang, M. J. Bremner, J. M. Martinis, and H. Neven, Characterizing quantum supremacy in near- term devices, Nature Physics14, 595 (2018)

    2018

  14. [14]

    Amico, R

    L. Amico, R. Fazio, A. Osterloh, and V. Vedral, En- tanglement in many-body systems, Reviews of Modern Physics80, 517 (2008)

    work page 2008

  15. [15]

    Poulin, A

    D. Poulin, A. Qarry, R. Somma, and F. Verstraete, Quan- tum Simulation of Time-Dependent Hamiltonians and the Convenient Illusion of Hilbert Space, Physical Re- view Letters106, 170501 (2011)

    work page 2011

  16. [16]

    Vidal, Efficient Classical Simulation of Slightly Entan- gled Quantum Computations, Physical Review Letters 91, 147902 (2003)

    G. Vidal, Efficient Classical Simulation of Slightly Entan- gled Quantum Computations, Physical Review Letters 91, 147902 (2003)

    work page 2003

  17. [17]

    Schollw¨ ock, The density-matrix renormalization group in the age of matrix product states, Annals of Physics 326, 96 (2011)

    U. Schollw¨ ock, The density-matrix renormalization group in the age of matrix product states, Annals of Physics 326, 96 (2011)

    2011

  18. [18]

    Ge and J

    Y. Ge and J. Eisert, Area laws and efficient descriptions of quantum many-body states, New Journal of Physics 18, 083026 (2016)

    work page 2016

  19. [19]

    Haferkamp, D

    J. Haferkamp, D. Hangleiter, J. Eisert, and M. Gluza, Contracting projected entangled pair states is average- case hard, Phys. Rev. Res.2, 013010 (2020)

    2020

  20. [20]

    Haghshenas, M

    R. Haghshenas, M. J. O’Rourke, and G. K.-L. Chan, Con- version of projected entangled pair states into a canonical form, Phys. Rev. B100, 054404 (2019)

    work page 2019

  21. [21]

    Evenbly, Gauge fixing, canonical forms, and optimal truncations in tensor networks with closed loops, Phys

    G. Evenbly, Gauge fixing, canonical forms, and optimal truncations in tensor networks with closed loops, Phys. Rev. B98, 085155 (2018)

    2018

  22. [22]

    Tindall and M

    J. Tindall and M. Fishman, Gauging tensor networks with belief propagation, SciPost Physics15, 222 (2023)

    2023

  23. [23]

    H. C. Jiang, Z. Y. Weng, and T. Xiang, Accurate De- termination of Tensor Network State of Quantum Lat- tice Models in Two Dimensions, Physical Review Letters 101, 090603 (2008)

    work page 2008

  24. [24]

    V. Murg, F. Verstraete, and J. I. Cirac, Variational study of hard-core bosons in a two-dimensional optical lattice 7 using projected entangled pair states, Phys. Rev. A75, 033605 (2007)

    work page 2007

  25. [25]

    Or´ us and G

    R. Or´ us and G. Vidal, Simulation of two-dimensional quantum systems on an infinite lattice revisited: Cor- ner transfer matrix for tensor contraction, Phys. Rev. B 80, 094403 (2009)

    2009

  26. [26]

    Z. Y. Xie, H. C. Jiang, Q. N. Chen, Z. Y. Weng, and T. Xiang, Second renormalization of tensor-network states, Phys. Rev. Lett.103, 160601 (2009)

    2009

  27. [27]

    Dziarmaga, Time evolution of an infinite projected en- tangled pair state: Neighborhood tensor update, Physical Review B104, 094411 (2021)

    J. Dziarmaga, Time evolution of an infinite projected en- tangled pair state: Neighborhood tensor update, Physical Review B104, 094411 (2021)

    work page 2021

  28. [28]

    Haegeman, J

    J. Haegeman, J. I. Cirac, T. J. Osborne, I. Pi ˚Aorn, H. Verschelde, and F. Verstraete, Time-dependent vari- ational principle for quantum lattices, Physical Review Letters107, 1 (2011)

    work page 2011

  29. [29]

    Haegeman, C

    J. Haegeman, C. Lubich, I. Oseledets, B. Vandereycken, and F. Verstraete, Unifying time evolution and optimiza- tion with matrix product states, Physical Review B94, 165116 (2016)

    2016

  30. [30]

    Alkabetz and I

    R. Alkabetz and I. Arad, Tensor networks contraction and the belief propagation algorithm, Physical Review Research3, 023073 (2021)

    2021

  31. [31]

    Evenbly, N

    G. Evenbly, N. Pancotti, A. Milsted, J. Gray, and G. K.- L. Chan, Loop Series Expansions for Tensor Networks (2024)

    work page 2024

  32. [32]

    A. D. King, J. Raymond, T. Lanting, R. Harris, A. Zucca, et al., Quantum critical dynamics in a 5,000-qubit pro- grammable spin glass, Nature617, 61 (2023)

    work page 2023

  33. [33]

    Bernaschi, I

    M. Bernaschi, I. Gonz´ alez-Adalid Pemart ´ ın, V. Mart ´ ın- Mayor, and G. Parisi, The quantum transition of the two- dimensional Ising spin glass, Nature631, 749 (2024)

    work page 2024

  34. [34]

    Quantum dynamics in frustrated Ising fullerenes

    A. Lopez-Bezanilla, W. Bernoudy, K. Boothby, J. Ray- mond, A. Nocera, and A. D. King, Quantum dynam- ics in frustrated Ising fullerenes (2025), arXiv:2505.08994 [quant-ph]

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  35. [35]

    Tindall, Personal communication (2025)

    J. Tindall, Personal communication (2025)

    work page 2025

  36. [36]

    Chern, K

    K. Chern, K. Boothby, J. Raymond, P. Farr´ e, and A. D. King, Tutorial: Calibration refinement in quan- tum annealing, Frontiers in Computer Science5, 1238988 (2023)

    work page 2023

  37. [37]

    Tindall and M

    J. Tindall and M. S. Rudolph, TensorNetworkQuantum- Simulator (2025)

    work page 2025

  38. [38]

    iTensorNetworks.jl (2023)

    work page 2023

  39. [39]

    bimodal” model, or the uniform distribution over the values [−1,1], discretized to a step size of 1 256 and excepting 0, in the “uniform

    M. Fishman, S. White, and E. Stoudenmire, The ITen- sor Software Library for Tensor Network Calculations, SciPost Physics Codebases4, 4 (2022). ACKNOWLEDGMENTS The authors are grateful to Joseph Tindall for help- ful discussions. Work at UBC was supported by Natural Sciences and Engineering Research Council of Canada (NSERC) Alliance Quantum Program (Gran...

    work page 2022

This paper was first reviewed by deepseek-v4-flash on August 5, 2026.