Pith. sign in

REVIEW 4 major objections 4 minor 49 references

Forecasting Quantum Observables: A Compressed Sensing Approach with Performance Guarantees

T0 review · 4 major / 4 minor · reviewed 2026-08-04 · deepseek-v4-flash

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

Pith's one-line read A certification test based on atomic norm minimization can decide whether a spectral model of a quantum observable – its Bohr frequencies and amplitudes – truly matches the unitary time evolution, and certified models keep forecasting error

desk verdict Useful filtering heuristic and honest empirical work, but the title's 'performance guarantees' outrun what the certificate actually proves: it checks consistency with the training window, not the full unitary evolution. read the letter →

arxiv 2510.14897 v3 pith:RIHGY3F5 submitted 2025-10-16 quant-ph

Víctor Valls , Albert Akhriev , Olatz Sanz Larrarte , Javier Oliva del Moral , Štěpán Šmíd , Josu Etxezarreta Martinez , Sergiy Zhuk , Dmytro Mishagli This is my paper
classification quant-ph
keywords atomicnormminimizationquantumobservableforecastingspectralcertificationdualpolynomialBohrfrequenciescompressedsensingspinchainsunitarytimeevolution
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

Data-driven forecasting of quantum observables has been empirical: algorithms like ESPRIT and DMD fit few sinusoids to early measurements, but no one knew whether the fitted model was the actual dynamics or just a good fit. The paper asks whether a given spectral model can be certified as consistent with unitary time evolution, and answers with a test built from atomic norm minimization. The test uses the dual polynomial: if it touches ±1 exactly at the candidate frequencies and stays below 1 everywhere else, with a small duality gap and well-separated frequencies, the model is certified. Across 5,850 spin-chain time evolutions, certified models forecast the remaining measurements with average error below 0.1 in 97% of cases and below 0.05 in 91–99% of cases, even with shot noise. The framework works for any forecasting algorithm, turning extrapolation from a heuristic into a decision procedure whenever the dynamics are sparse.

What carries the argument

The central object is the dual polynomial Q(f) = ⟨a(f), q⟩ of the atomic norm minimization dual problem, where a(f) is the atom vector of a continuous-frequency sinusoid sampled at the measurement times. The certification pipeline solves an approximate dual problem on a fine frequency grid, then checks three criteria: the duality gap between the primal and dual SDP values, the minimum separation of candidate frequencies (≥ 4/m), and the shape of the dual polynomial (|Q(f)| ≈ 1 at candidate frequencies, <1 elsewhere). The atomic norm itself – the continuous analogue of the ℓ1 norm – is the machinery that enforces sparsity in the frequency domain and gives the uniqueness guarantee whenever the

What would settle it

The paper itself supplies a falsifier in Appendix C2: for the correlator ⟨Z1(t)Z30(t)⟩ in a 30-site transverse-field Ising model, a model certified on the first 60 measurements forecasts a constant, missing the oscillations that arrive after the light cone. Any case where certification passes but the post-window dynamics contain new frequencies would refute the claim that certification guarantees forecasting accuracy; the test is to run the pipeline on such a signal and observe the forecast error exceeding the tolerance.

Watch

Extended reading notes

Core claim

The central claim is that atomic norm minimization provides a certificate of consistency for any candidate spectral model of a unitary quantum time evolution. When the true signal is a sparse sum of Bohr frequencies separated by at least 4/m, the dual polynomial of the ANM problem is a uniqueness certificate: it attains modulus one precisely at the true frequencies and is strictly below one otherwise. The paper proposes checking three conditions on a candidate model – a small duality gap, frequency separation, and the dual-polynomial condition – and shows numerically that certified models, obtained from any forecasting algorithm, yield average forecasting error below 0.1 in 97% of cases and

Load-bearing premise

The certificate computed from the training window is assumed to certify the candidate model's consistency with the full-time signal, which requires that every Bohr frequency that becomes relevant during the forecasting horizon already has non-negligible amplitude in the training window.

Editorial extensions

If this is right

  • Any forecasting algorithm's spectral output can be certified or rejected without knowledge of the true Hamiltonian, giving a formal performance guarantee for extrapolation.
  • The framework applies to systems beyond exact diagonalization, as demonstrated on a 100-qubit tilted Ising model, provided the dynamics are sparse and frequencies are well separated.
  • Certified models keep average forecasting error below 0.1 in 97% of cases for noiseless data, and remain accurate at the same threshold under shot noise, making the certificate a practical validation tool for near-term simulators.
  • Successful certification requires sufficiently many early-time measurements; the empirical evidence suggests a threshold around 30 measurements for reliable certification in the tested regimes.
  • If certification fails, the forecasting model should not be trusted, so the pipeline converts forecasting from an ungrounded heuristic into a decision process with a stated condition for its validity.

Reading between the lines

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

  • The certificate's authority is bounded by the training window's information content: the authors' own light-cone example shows a model certified on the first 60 samples of a correlator misses oscillations that appear later, meaning a certificate does not guarantee prediction of phases not already present in the data.
  • The dual polynomial's shape away from the candidate frequencies could be used as a quantitative probe for 'missing frequencies' – if the polynomial is far below 1 on a wide band, the candidate model may be underfitting, suggesting a diagnostic beyond the binary pass/fail.
  • Because the certificate is algorithm-agnostic, it could serve as a model-selection rule: given multiple candidate spectral models from different forecasting methods, pick the first one that certifies, or use certification speed as a criterion for choosing among them.
  • The framework could be combined with sparse-Pauli or classical-shadow data generation to vet spectral models for systems with hundreds of qubits, extending the 100-site demonstration to regimes where direct state-vector simulation is intractable.
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 / 4 minor

Summary. The paper proposes an atomic-norm-minimization (ANM) framework for certifying spectral models of quantum-observable time series. Given a candidate Bohr-frequency/amplitude model produced by any forecasting algorithm, the method solves the ANM dual problem approximately, evaluates three criteria — duality gap, frequency separation ≥ 4/m, and dual-polynomial floor at the candidate frequencies — and declares the model certified if empirically chosen thresholds are met. The authors claim that certified models are consistent with unitary quantum time evolution and validate this on spin-chain Hamiltonians (8–20 sites, 5850 TEs, 5 algorithms), reporting high conditional reliability (e.g., >97% below error 0.1 for exact TE in Table I). The paper also includes a 100-site tilted-Ising example and a light-cone counterexample in Appendix C2.

Significance. If the certification were rigorous, the framework would be valuable: it would give a model-agnostic, data-driven validation tool for quantum-observable extrapolation, with potential relevance to near-term quantum hardware. The paper’s strengths are the breadth of empirical testing (five forecasting algorithms, exact and shot-noise data, multiple Hamiltonians and observables), the use of exact diagonalization as an independent ground truth, and the publicly available code. However, the paper does not deliver a mathematical guarantee: the certificate is implemented as a discretized, Adam-solved heuristic with ad hoc thresholds, and the theoretical ANM conditions are not verified exactly. The central claims in the title and abstract therefore overstate what is established. A revised version that reframes the contribution as a heuristic validation test with explicit limitations would be of practical interest.

major comments (4)
  1. [Sec. III B, App. A.2; Eqs. (5)-(6)] The 'certificate' is not a certificate in the rigorous sense. The dual problem in Eq. (5) is infinite-dimensional, but the implementation replaces it by a 1000-point grid and an Adam solve with 500 iterations and penalty λ=15, then applies empirically chosen thresholds (|Q(f)|≥0.98, gap≤0.05 for exact TE; 0.95/0.5 for noisy). No error bound is given for the discretization error, optimization residual, or threshold effects. The cited ANM theorems ([11, Thm 1.2]; [12, Prop 2.4]) require exact conditions: Q(f)=1 on the true support, |Q(f)|<1 elsewhere, and separation >4/m. None of these are verified exactly. Thus the paper does not establish a 'performance guarantee'; at best it provides a heuristic validation test. The abstract and title should be revised to reflect this.
  2. [App. C2, Fig. 8b; Sec. IV B] The certificate is computed from the m training samples only and cannot rule out the appearance of new Bohr frequencies after the training window. The paper's own light-cone example (⟨Z1Z30⟩, 30-site TFIM) is the relevant test: the signal is essentially zero for the first 60 samples and oscillates later; an OMP model trained on the first 60 samples fails to capture the later oscillations. The certification outcome for this model is not reported. If it passed the criteria of Sec. III B, it is a direct counterexample to the unconditional reading of 'certified models yield ... error below 0.1'; if it failed, the example still demonstrates that certification is a consistency check of the training segment, not of the full-time unitary dynamics. Please report the certification status and explicitly qualify the reliability claims accordingly.
  3. [Eq. (6); App. A.2] The dual LP as written is mathematically ill-posed: a(f) in Eq. (3) is complex, so the constraint −1≤⟨a(f),q⟩≤1 in Eq. (6) is not a valid real inequality. If the implementation instead constrains |Q(f)|≤1 (or separately real and imaginary parts), this should be written out explicitly. The restriction q∈R^m (rather than q∈C^m) also needs justification. Because the duality gap and dual-polynomial floor are the basis of certification, an incorrect dual formulation would invalidate the procedure; the authors should clarify the actual formulation used.
  4. [Tables I-II; Fig. 4; abstract] Table I reports the fraction of certified runs whose forecasting error is below ϵ (conditional on certification), not the 'average forecasting error'. The abstract's wording 'average forecasting error below 0.1 ... in 97% of cases' conflates these. Moreover, Fig. 4 shows that at the first certification (the earliest k meeting the criteria), errors below 0.1 occur in only about 75% of cases for some exact-TE algorithms, so the 97% figure is an aggregate over all k and does not represent typical first-certification performance. The authors should report error distributions at first certification and state clearly that all reliability numbers are conditional on certification and on the specific test ensemble.
minor comments (4)
  1. [Sec. IV A] The dataset description lists 15 Hamiltonians × 5 initial states × 6 observables = 450 TEs, but 5850 TEs are reported. Please clarify the counting, e.g., the number of system sizes (8–20 gives 13 sizes) and boundary conditions.
  2. [Table I] Use 'Prony' consistently; the abbreviation 'PNY' is not defined in the caption.
  3. [Appendix A.2] Please report the initialization, stopping criterion, and randomness of the Adam solve, and whether results are averaged over seeds. Since the certification thresholds are empirical, solver variability directly affects the reliability statistics.
  4. [Sec. III B] The sentence 'the duality gap is considered admissible if this is below a prescribed tolerance' should state how the gap is computed from the approximate dual solution and how the tolerances were chosen (rather than only listing their values).

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the certification conditions are external ANM theorems and the headline claim is tested against exact diagonalization.

full rationale

The paper's derivation chain is: (1) each observable is a finite sum of Bohr harmonics (Eq. 1); (2) if the harmonic representation is sparse and well-separated, ANM's dual certificate uniquely identifies it (cited external theorems, [11,12]); (3) the pipeline checks the candidate model against three independent criteria that operationalize those theorems; (4) the empirical claim is then measured against exact-diagonalization ground truth. None of these steps defines the target in terms of the input. The dual polynomial is computed from the measurements y (Eq. 6), not from the candidate model; the duality-gap bound is a check between a candidate-based primal feasible point and a y-based dual lower bound; separation is checked on the candidate frequency set. The forecasting-error claim in Table I is an out-of-sample test, not a fitted quantity. The only model-selection element is the 'empirically chosen' certification thresholds, but these are tolerances for numerical solving, not parameters of the forecasting model, and they do not make the certified-model claim true by construction. The paper itself acknowledges the light-cone limitation (Appendix C2), which weakens the universality of the guarantee but is not circularity. No load-bearing self-citation appears; the ANM uniqueness results are imported from Candès–Fernandez-Granda and Tang et al.

Assumptions & free parameters 8 free parameters · 5 assumptions · 0 invented entities

The paper's central claim rests on the ANM uniqueness theory (external), on a domain assumption that the true TE is sparse in Bohr frequencies that are active in the observed window, and on an ad-hoc assumption that a 1000-point discretized dual check with hand-tuned thresholds certifies the continuous condition. The empirical thresholds are fitted to the data, not derived.

free parameters (8)
  • certification threshold (exact TE): dual polynomial floor = 0.98
    Empirically chosen; lower bound on |Q(f)| at candidate frequencies for certification on noiseless data. Directly affects which runs are certified and the reported reliability.
  • certification threshold (exact TE): duality gap = 0.05
    Empirically chosen upper bound for certification; controls stringency.
  • certification threshold (exact TE): amplitude cutoff = 0.025
    Frequencies with smaller amplitudes are ignored; chosen empirically.
  • certification threshold (shot noise): dual polynomial floor = 0.95
    Looser threshold for noisy data.
  • certification threshold (shot noise): duality gap = 0.5
    Looser gap threshold for noisy data.
  • certification threshold (shot noise): amplitude cutoff = 0.05
    Higher cutoff for noisy data.
  • dual frequency grid resolution = 1000 points
    The dual atomic-norm constraint is enforced on a 1000-point grid; finer grid gives better approximation.
  • dual solver (Adam) hyperparameters = λ=15, lr=0.1, 500 iterations
    Approximate dual solution; convergence and constraint satisfaction not formally certified.
assumptions (5)
  • standard math Expectation values of a closed quantum system are finite sums of harmonics (Eq. 1).
    Follows from spectral decomposition of e^{-iHτ}. Central to the formulation.
  • domain assumption The true TE is sparse in the Bohr-frequency basis and frequencies are separated by at least 4/m.
    Required for ANM uniqueness (Thm 1.2 in [11], Thm 1.1 in [12]); the paper assumes this holds for certified TEs.
  • domain assumption Frequencies active in the observed window dominate the future window (no late-time spectral support growth).
    Needed to justify that a sparse model fit to the observed m samples forecasts the future; contradicted by the light-cone example in Appendix C2.
  • ad hoc to paper Discretized dual check on 1000 grid points approximates the continuous constraint |Q(f)|≤1.
    The paper asserts the error is negligible for a sufficiently fine grid but gives no bound; certification relies on this.
  • domain assumption Shot noise model: 1000-shot projective measurements; loosened thresholds account for noise.
    The paper uses empirical threshold adjustments rather than a noise-aware ANM guarantee.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Forecasting Quantum Observables: A Compressed Sensing Approach with Performance Guarantees." pith.science (2026). https://pith.science/paper/RIHGY3F5

@misc{pith2026251014897,
  author       = {Pith},
  title        = {Pith review of: Forecasting Quantum Observables: A Compressed Sensing Approach with Performance Guarantees},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RIHGY3F5}},
  note         = {Machine review of arXiv:2510.14897}
}
abstract

Data-driven extrapolation methods aim to extend the dynamics of quantum observables from measurements, but they often lack guarantees on prediction accuracy. We introduce a framework based on atomic norm minimization that can certify whether the spectral model learned by a forecasting algorithm -- i.e., Bohr frequencies and amplitudes -- is consistent with unitary quantum time evolution. Certification holds when the dynamics are governed by a small number of well-separated Bohr frequencies. We validate the approach on multiple forecasting algorithms applied to spin-chain Hamiltonians with 8--20 sites. Comparing with exact diagonalization, certified models yield an average forecasting error below 0.1 (observable range $[-1, 1]$) in 97\% of cases and below 0.05 in 91--99\% of cases. Even in the presence of noise, certified models remain robust at the 0.1 error threshold.

Figures

Figures reproduced from arXiv: 2510.14897 by the authors.

Figure 1
Figure 1. FIG. 1 [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 3
Figure 3. An algorithm (e.g., ESPRIT [5]) takes as in￾put a vector y with m measurements and returns candi￾date frequencies ∑ Ω and amplitudes c(f) such that yk ≈ f∈Ω c(f) e i2πf τk , k ∈ {1, . . . , m}. These candidates fre￾quencies and amplitudes are then certified by solving the ANM dual problem in Eq. (5). Certificates. To certify the model, we evaluate three key criteria: the duality gap, frequency separation, and the du… view at source ↗
Figure 4
Figure 4. FIG. 4 [PITH_FULL_IMAGE:figures/full_fig_p004_4.png] view at source ↗
Figures from the paper (5 more)
Figure 5
Figure 5. Figure 5: FIG. 5 [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]
Figure 6
Figure 6. Figure 6: FIG. 6 [PITH_FULL_IMAGE:figures/full_fig_p010_6.png]
Figure 7
Figure 7. Figure 7: FIG. 7 [PITH_FULL_IMAGE:figures/full_fig_p011_7.png]
Figure 8
Figure 8. Figure 8: FIG. 8 [PITH_FULL_IMAGE:figures/full_fig_p011_8.png]
Figure 9
Figure 9. Figure 9: shows the time it takes MOSEK [45] (a commercial SDP solver) to solve the problem in Eq. (4). The SDP is run on an Intel i7-14700 CPU with 32 GB of memory. 0 50 100 150 200 250 0 50 100 150 200 250 300 350 400 Time (s) Number of measurements FIG. 9. Time to solve the S…

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

49 extracted references · 1 canonical work pages

  1. [1]

    Fauseweh, Quantum many-body simulations on digi- tal quantum computers: State-of-the-art and future chal- lenges, Nature Communications 15, 2123 (2024)

    B. Fauseweh, Quantum many-body simulations on digi- tal quantum computers: State-of-the-art and future chal- lenges, Nature Communications 15, 2123 (2024)

    2024

  2. [2]

    Y. Kim, A. Eddins, S. Anand, K. X. Wei, E. van den Berg, S. Rosenblatt, H. Nayfeh, Y. Wu, M. Zaletel, K. Temme, and A. Kandala, Evidence for the utility of quantum computing before fault tolerance, Nature 618, 500 (2023)

    2023

  3. [3]

    E. D. Switzer, N. Robertson, N. Keenan, Á. Rodríguez, A. D’Urbano, B. Pokharel, T. S. Rahman, O. Shtanko, S. Zhuk, and N. Lorente, Realization of Two-dimensional Discrete Time Crystals with Anisotropic Heisenberg Coupling (2025), arXiv:2501.18036 [quant-ph]

    arXiv 2025

  4. [4]

    Temme, S

    K. Temme, S. Bravyi, and J. M. Gambetta, Error miti- gation for short-depth quantum circuits, Physical review letters 119, 180509 (2017)

    2017

  5. [5]

    Erpenbeck, Y

    A. Erpenbeck, Y. Zhu, Y. Yu, L. Zhang, R. Gerum, O. Goulko, C. Yang, G. Cohen, and E. Gull, Com- pact representation and long-time extrapolation of real-time data for quantum systems, arXiv preprint arXiv:2506.13760 (2025)

    arXiv 2025

  6. [6]

    Manos, M

    K. Manos, M. Weilenmann, and M. Navascues, Ex- trapolation of quantum measurement data (2025), arXiv:2507.06912

    arXiv 2025

  7. [7]

    Kaneko, M

    R. Kaneko, M. Imada, Y. Kabashima, and T. Ohtsuki, Forecasting long-time dynamics in quantum many-body systems by dynamic mode decomposition, Physical Re- view Research 7, 013085 (2025)

    2025

  8. [8]

    Y. Shen, A. Buzali, H.-Y. Hu, K. Klymko, D. Camps, S. F. Yelin, and R. Van Beeumen, Efficient measurement- driven eigenenergy estimation with classical shadows, arXiv preprint arXiv:2409.13691 (2024). 6

    arXiv 2024

Show all 49 references
  1. [9]

    B. N. Bhaskar, G. Tang, and B. Recht, Atomic norm denoising with applications to line spectral estimation, IEEE Transactions on Signal Processing 61, 5987 (2013)

    2013

  2. [10]

    Chandrasekaran, B

    V. Chandrasekaran, B. Recht, P. A. Parrilo, and A. S. Willsky, The convex geometry of linear inverse prob- lems, Foundations of Computational mathematics 12, 805 (2012)

    2012

  3. [11]

    E. J. Candès and C. Fernandez-Granda, Towards a math- ematical theory of super-resolution, Communications on pure and applied Mathematics 67, 906 (2014)

    2014

  4. [12]

    G. Tang, B. N. Bhaskar, P. Shah, and B. Recht, Com- pressed sensing off the grid, IEEE transactions on infor- mation theory 59, 7465 (2013)

    2013

  5. [13]

    Plonka and M

    G. Plonka and M. Tasche, Prony methods for recov- ery of structured functions, GAMM-Mitteilungen 37, 239 (2014)

    2014

  6. [14]

    S. R. White and D. A. Huse, Numerical renormalization- group study of low-lying eigenstates of the antiferromag- netic s=1 heisenberg chain, Phys. Rev. B 48, 3844 (1993)

    1993

  7. [15]

    Honecker and Wessel, Magnetocaloric effect in quan- tum spin-s chains, Condensed Matter Physics 12, 399410 (2009)

    2009

  8. [16]

    Giamarchi, Quantum Physics in One Dimension (Ox- ford University Press, 2003)

    T. Giamarchi, Quantum Physics in One Dimension (Ox- ford University Press, 2003)

    2003

  9. [17]

    H. J. Schulz, Phase diagrams and correlation exponents for quantum spin chains of arbitrary spin quantum num- ber, Phys. Rev. B 34, 6372 (1986)

    1986

  10. [18]

    Affleck, Quantum spin chains and the haldane gap, Journal of Physics: Condensed Matter 1, 3047 (1989)

    I. Affleck, Quantum spin chains and the haldane gap, Journal of Physics: Condensed Matter 1, 3047 (1989)

    1989

  11. [19]

    den Nijs and K

    M. den Nijs and K. Rommelse, Preroughening transitions in crystal surfaces and valence-bond phases in quantum spin chains, Phys. Rev. B 40, 4709 (1989)

    1989

  12. [20]

    Kurmann, H

    J. Kurmann, H. Thomas, and G. Müller, Antiferromag- netic long-range order in the anisotropic quantum spin chain, Physica A: Statistical Mechanics and its Applica- tions 112, 235 (1982)

    1982

  13. [21]

    A. K. Bera, B. Lake, A. T. M. N. Islam, B. Klemke, E. Faulhaber, and J. M. Law, Field-induced mag- netic ordering and single-ion anisotropy in the quasi- one-dimensional haldane chain compound srni 2v2o8: A single-crystal investigation, Phys. Rev. B 87, 224423 (2013)

    2013

  14. [22]

    Nachtergaele, W

    B. Nachtergaele, W. Spitzer, and S. Starr, Ferromagnetic ordering of energy levels, Journal of Statistical Physics 116, 719738 (2004)

    2004

  15. [23]

    S. M. Giampaolo, G. Adesso, and F. Illuminati, Theory of ground state factorization in quantum cooperative sys- tems, Phys. Rev. Lett. 100, 197201 (2008)

    2008

  16. [24]

    J. R. Johansson, P. D. Nation, and F. Nori, Qutip: An open-source python framework for the dynamics of open quantum systems, Computer physics communica- tions 183, 1760 (2012)

    2012

  17. [25]

    Sauer, Pronys method: an old trick for new prob- lems, Snapshots of modern mathematics from Oberwol- fach (2018)

    T. Sauer, Pronys method: an old trick for new prob- lems, Snapshots of modern mathematics from Oberwol- fach (2018)

    2018

  18. [26]

    Chen and D

    S. Chen and D. Donoho, Basis pursuit, in Proceedings of 1994 28th Asilomar Conference on Signals, Systems and Computers, Vol. 1 (IEEE, 1994) pp. 41–44

    1994

  19. [27]

    Locatello, R

    F. Locatello, R. Khanna, M. Tschannen, and M. Jaggi, A Unified Optimization View on Generalized Match- ing Pursuit and Frank-Wolfe, in Proceedings of the 20th International Conference on Artificial Intelligence and Statistics, Proceedings of Machine Learning Research, Vol. 54, e...

    2017

  20. [28]

    Beck and M

    A. Beck and M. Teboulle, A fast iterative shrinkage- thresholding algorithm for linear inverse problems, SIAM journal on imaging sciences 2, 183 (2009)

    2009

  21. [29]

    Paulraj, R

    A. Paulraj, R. Roy, and T. Kailath, Estimation Of Signal Parameters Via Rotational Invariance Techniques- Es- prit, in Nineteeth Asilomar Conference on Circuits, Sys- tems and Computers, 1985. (1985) pp. 83–89

    1985

  22. [30]

    Roy and T

    R. Roy and T. Kailath, Esprit-estimation of signal parameters via rotational invariance techniques, IEEE Transactions on Acoustics, Speech, and Signal Processing 37, 984 (1989)

    1989

  23. [31]

    https://github.com/victor-ibm/ quantum-observable-forecasting

  24. [32]

    It is sufficient to consider [0, 1 2 ] instead of [− 1 2, 1 2 ] since the frequency spectrum is symmetric

  25. [33]

    Stoica, R

    P. Stoica, R. L. Moses, et al. , Spectral analysis of signals , Vol. 452 (Pearson Prentice Hall Upper Saddle River, NJ, 2005)

    2005

  26. [34]

    Hua and T

    Y. Hua and T. Sarkar, Matrix pencil method for estimat- ing parameters of exponentially damped/undamped sinu- soids in noise, IEEE Transactions on Acoustics, Speech, and Signal Processing 38, 814 (1990)

    1990

  27. [35]

    Wax and T

    M. Wax and T. Kailath, Detection of signals by informa- tion theoretic criteria, IEEE Transactions on Acoustics, Speech, and Signal Processing 33, 387 (1985)

    1985

  28. [36]

    Golyandina and N

    N. Golyandina and N. Zvonarev, Information criteria for signal extraction using singular spectrum analysis: White and red noise, Algorithms 17, 10.3390/a17090395 (2024)

    2024 doi

  29. [37]

    Golyandina, V

    N. Golyandina, V. Nekrutkin, and A. A. Zhigljavsky, Analysis of time series structure - ssa and related tech- niques, in Monographs on statistics and applied probabil- ity (2001)

    2001

  30. [38]

    D. P. Kingma and J. Ba, Adam: A method for stochastic optimization, arXiv preprint arXiv:1412.6980 (2014)

    2014 arXiv

  31. [39]

    Karthik, A

    J. Karthik, A. Sharma, and A. Lakshminarayan, Entan- glement, avoided crossings, and quantum chaos in an ising model with a tilted magnetic field, Phys. Rev. A 75, 022304 (2007)

    2007

  32. [40]

    Y. Liu, M. Li, J. Bao, B. Guo, and Z. Sun, Multipartite nonlocality in an ising model with a tilted magnetic field at zero and finite temperatures, Physics Letters A 450, 128396 (2022)

    2022

  33. [41]

    Dutta, G

    A. Dutta, G. Aeppli, B. K. Chakrabarti, U. Divakaran, T. F. Rosenbaum, and D. Sen, Quantum Phase Transi- tions in Transverse Field Spin Models: From Statistical Physics to Quantum Information (Cambridge University Press, 2015)

    2015

  34. [42]

    M. S. Rudolph, T. Jones, Y. Teng, A. Angrisani, and Z. Holmes, Pauli Propagation: A Computa- tional Framework for Simulating Quantum Systems, https://arxiv.org/abs/2505.21606v1 (2025)

    2025 arXiv

  35. [43]

    Begušić and G

    T. Begušić and G. K.-L. Chan, Real-time operator evo- lution in two and three dimensions via sparse pauli dy- namics, PRX Quantum 6, 020302 (2025)

    2025

  36. [44]

    (Anthony) Chen, A

    C.-F. (Anthony) Chen, A. Lucas, and C. Yin, Speed lim- its and locality in many-body quantum dynamics, Rep. Prog. Phys. 86, 116001 (2023)

    2023

  37. [45]

    ApS, MOSEK Optimization Software (2025), version 11.0

    M. ApS, MOSEK Optimization Software (2025), version 11.0. 7 Appendix A: Algorithms

    2025

  38. [46]

    complex (analytic), stationary, and ergodic Gaussian random processes with zero mean and a positive-definite covariance matrix

    Algorithm for finding frequencies and amplitudes a. Prony’s method Pronys method is implemented as described in [ 25]. The method requires that the number of measurements m satisfies ≥ 2|Ω|, where |Ω| denotes the number of frequencies. However, |Ω| is typically unknown. To addre...

  39. [47]

    periodic

    Algorithm to solve the dual problem in Eq. (5) We solve the problem in Eq. ( 5) with Adam [ 38]. The objective function is f (q) = ⟨y,q ⟩ +λ L∑ j=1 max{0, |(Aq)j| − 1}2, (A2) where y ∈ Rm is the vector of measurements, and A ∈ RL×m a matrix with rows equal to the atoms in Eq. ...

  40. [48]

    A natural question is whether this pipeline can scale up to systems with hundreds of sites

    Ising model with a tilted field and potential scalability In the main text, we extensively tested the proposed certification pipeline for Hamiltonians involving 8–20 spin sites. A natural question is whether this pipeline can scale up to systems with hundreds of sites. Our main ...

  41. [49]

    light cone

    T wo-point correlator example: invalid certification Quantum information spreads in spin chains with a finite speed, defined by the Lieb–Robinson bound [ 44]. Fig. 8a shows the “light cone” heatmap produced by a two-point correlator, ⟨Z1(t)Zj(t)⟩ with j = 2, 3,...,n , for a TFIM ...

Pith tools

Reviewed August 4, 2026 · model on record in the stance chip above.