Pith. sign in

REVIEW 4 major objections 6 minor 101 references

Input-Output Optics as a Causal Time Series Mapping: A Generative Machine Learning Solution

T0 review · 4 major / 6 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read A temporal convolutional autoencoder's minimal latent dimension ranks the complexity of optical input–output dynamics in driven Ising models.

desk verdict A plausible ML demo with an overreaching complexity measure: the TCN mapping works, but the latent-dimension ranking is not yet established. read the letter →

arxiv 2411.19897 v2 pith:IYM2PRW7 submitted 2024-11-29 quant-ph physics.optics

Abhijit Sen , Bikram Keshari Parida , Kurt Jacobs , Denys I. Bondar This is my paper · ORCID
classification quant-phphysics.optics
keywords temporalconvolutionalnetworksvariationalautoencoderquantumIsingmodelinput-outputtimeserieslatentspacecomplexityamplitude-awarepermutationentropydynamicssimulationmachinelearningforoptics
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 treats light–matter interaction as a causal mapping from an input pulse time series to an output magnetization time series, and asks whether a neural network can learn that mapping directly from data. Using transverse and non-integrable Ising chains driven by magnetic-field pulses, the authors show that temporal convolutional network (TCN) autoencoders can reproduce the input–output relation with high accuracy. Their central proposal is that the dimension of the smallest latent space that still yields an accurate model is a measure of the complexity of the underlying quantum dynamics. The paper also reports that a variational autoencoder, a generative model, outperforms a plain autoencoder, most clearly on the most complex case, and that the latent-space ranking agrees with amplitude-aware permutation entropy.

What carries the argument

The central object is the autoencoder temporal convolutional network, a causal encoder–decoder in which dilated convolutions with kernel size 3 and dilation factors (2,4,8) compress each 512-step input pulse into a latent bottleneck and decode it into the output pulse. The latent dimension is the number of neurons at the bottleneck, and the paper's complexity measure is the smallest such dimension that still yields a stable accurate model under the stated threshold. The variational variant replaces the deterministic bottleneck with a Gaussian latent distribution regularized by KL divergence, which is what makes it a generative model.

What would settle it

Recompute the minimal stable latent dimension for the same data sets under a stricter or looser accuracy threshold (for example, requiring 95% of test $R^2$ values above 0.95 instead of 85% above 0.85), or with a different encoder family such as an LSTM autoencoder, and check whether the relative ranking of the five cases changes. If the ranking changes, the proposed complexity measure is threshold- or architecture-dependent.

Watch

Extended reading notes

Core claim

The discovery is that the minimal viable latent-space dimension of a causal TCN autoencoder is a good complexity measure for quantum input–output dynamics. Across five driven-Ising data sets, the paper identifies, by training ten models per architecture, the smallest encoder–decoder configuration whose test-set $R^2$ values stay above the stability threshold, and finds that deeper and wider bottlenecks are required as the driving amplitude grows and as the Hamiltonian becomes non-integrable. Case 2, the strongly driven transverse Ising model, demands the largest minimal latent space and is ranked the most complex. The ranking matches the ordering given by amplitude-aware permutation entropy of the output time series, which the paper presents as independent corroboration. The paper further claims that a variational TCN autoencoder reaches lower loss and higher stable $R^2$ fractions than the deterministic autoencoder at comparable parameter counts, with over 90% of test inputs predicted below 10% error in the hardest case.

Load-bearing premise

The paper assumes that the smallest latent dimension found under its chosen accuracy threshold and manually selected architectures is an intrinsic complexity property of the dynamics, not an artifact of the threshold or the network family.

Editorial extensions

If this is right

  • Driven Ising dynamics can be simulated by a trained TCN autoencoder, predicting output magnetization pulses from unseen input pulses without solving the Schrödinger equation.
  • The minimal stable latent dimension ranks the five cases in an order consistent with amplitude-aware permutation entropy, so the measure can serve as a data-driven complexity diagnostic for regimes where perturbative orders are undefined.
  • The variational autoencoder's advantage indicates that generative, distribution-based encoders are preferable for learning strong-field or non-integrable responses.
  • The methodology extends in principle to any causal optical input–output relation, because the TCN architecture itself enforces causality rather than relying on an externally imposed causal model.

Reading between the lines

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

  • A threshold- and architecture-independence test is the natural next step: recomputing the minimal latent dimension at several accuracy thresholds and with a recurrent or transformer encoder would show whether the ranking is intrinsic to the dynamics or an artifact of the chosen network family.
  • Because the paper ties latent dimension to entropic complexity, one could further probe whether the minimal latent dimension tracks other information-theoretic quantities such as the quantum mutual information between input and output.
  • The VAE's probabilistic latent space could also yield calibrated uncertainty estimates for predicted output pulses, which the paper does not explore but which would be practically valuable in experimental settings.
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 / 6 minor

Summary. The paper treats the input-output relation h(t) → Y(t) of driven Ising models as a causal time-series mapping and trains temporal convolutional network (TCN) autoencoders on numerically generated data for five cases (transverse and non-integrable Ising chains at different driving amplitudes). The main methodological claim is that the dimension of the smallest latent space able to model the mapping accurately is a good complexity measure for the underlying quantum dynamics; this is supported by comparing TCN autoencoder architectures and by a comparison to amplitude-aware permutation entropy. The paper further claims that a variational autoencoder (VAE) significantly outperforms the traditional autoencoder on the most complex case (Case 2). Results are reported as R² distributions on test sets, with stability defined by a threshold criterion.

Significance. If the latent-space complexity measure were rigorously established, it would provide a data-driven, non-perturbative way to rank the complexity of input-output quantum dynamics, potentially useful for strong-field and many-body problems where perturbation theory fails. The paper ships code, uses a causal TCN architecture appropriate for time series, and provides a concrete comparison with an established entropy measure. However, the central measure is currently supported only by a small, hand-picked architecture search and a threshold-dependent stability criterion, so the paper's main conceptual contribution is not yet demonstrated to the standard required for publication.

major comments (4)
  1. [Sec. III C and Appendix B] The central claim that the reported architectures are the smallest latent spaces that accurately model the mapping is not established for Cases 2–5. Appendix B performs a minimality search only for Case 1, and even that search examines only four architectures along a single path (5-5-3, 5-5-2, 4-4-3, 3-3-2). For Cases 2–5, no smaller latent dimension is reported; for example, Case 2 is simply declared to have minimum architecture (12-12-10-10) without showing that (10-10-8-8) or (8-8-6-6) fails the stability criterion. The latent-dimension ranking (3, 10, 4, 6, 4) may therefore be an artifact of where the manual search stopped rather than a property of the quantum dynamics.
  2. [Sec. III B and Appendix B] The stability threshold is arbitrary and load-bearing. The criterion “more than 85% of test R² above 0.85” is used to define the minimum stable architecture, but Appendix B evaluates the candidate architectures using different thresholds (percentage above 0.98, 0.95, 0.90). For instance, the (5-5-3) architecture has one run with only 82% of R² above 0.98 (Table III), which would fail a stricter threshold such as “95% above 0.95.” The paper does not show that the latent-dimension ranking is robust to the choice of threshold, nor that the “instability” observed in Tables IV–VI is statistically meaningful beyond single-run fluctuations. Since the complexity measure is defined through this threshold, its threshold dependence must be characterized.
  3. [Sec. III E and Fig. 10] The permutation-entropy validation is confounded by input amplitude. The amplitude-aware permutation entropy Pnorm is computed on the output time series, and Fig. 10 shows that Pnorm increases monotonically with input amplitude for the transverse Ising model. The five studied cases differ not only in Hamiltonian and dynamics but also in driving amplitude (A = 1, 10, 1.5, 2.5, and a range 1–10), so the agreement between latent dimension and Pnorm may simply reflect that both quantities increase with driving strength rather than with intrinsic dynamical complexity. To support the claim that latent size tracks complexity, the authors should compare cases at matched amplitudes or demonstrate that the latent-size ranking persists after controlling for amplitude.
  4. [Sec. III D and Appendix C] The claim that the VAE “significantly outperforms” the traditional autoencoder is not statistically substantiated. In Appendix C, Table VII reports average percentages of R² above 0.90 as 96.7% for the VAE, 91.2% for TCN 1, and 93.2% for TCN 2, but the per-run values overlap substantially (e.g., VAE runs 4 and 5 at 91% and 90% versus TCN 2 runs 4, 5, and 10 at 97%, 95%, and 80%). The paper provides no significance test or confidence intervals, and the VAE has more than twice the parameters of the minimal TCN architecture, so the performance difference may be attributable to capacity rather than to the generative formulation. A statistical comparison over more runs or a matched-capacity baseline is needed.
minor comments (6)
  1. [Author affiliation] The affiliation line contains a typo: “Luisiana” should be “Louisiana.”
  2. [Sec. III C] There is a duplicated phrase “present the results in in III D”; remove the extra “in.”
  3. [Fig. 3 caption] The caption says “Deep Neural Network Autoencoder Architechture”; “Architechture” should be “Architecture.”
  4. [Sec. III C, Case 2] The sentence “The need for a deeper architecture suggests that the dynamics in Case 2 are more non-linear then in Case 1” uses “then” where “than” is intended.
  5. [Sec. III C, Case 2] The stability sentence “above 85% of {R2_i} lie above are above 0.90” contains a grammatical error; delete “lie above are.”
  6. [Appendix C] The parameter count of the VAE-TCN is inconsistent: Sec. III D and Table II state 34,127 trainable parameters, while Appendix C refers to “32,127” in two places. Please correct the inconsistency.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the TCN mapping is trained and evaluated on held-out test data, and the amplitude-aware permutation entropy comparison is an independent benchmark; the undersupported minimality claim is an evidentiary gap, not a circular reduction.

full rationale

The paper's central derivation chain is empirical rather than deductive: TCN autoencoders are trained on input-output pairs {h(i)->Y(i)} generated by solving the Schrödinger equation, and predictive quality is quantified by R^2 on held-out test sets (Eq. 6). The latent dimension is selected by an architecture search (Appendix B for Case 1; asserted for other cases), not by fitting to the permutation-entropy measure. The complexity validation in Sec. III E uses amplitude-aware permutation entropy computed from the output time series via the external Entropy Hub package; this is an independent, concurrently applied benchmark, and the paper does not derive the entropy values from the latent dimensions or vice versa. The self-citations in the introduction (Refs. [53-55]) concern quantum tracking control and are background material, not load-bearing for the complexity measure. Therefore no equation or fitted parameter is renamed as a prediction by construction. The identified weaknesses—the arbitrary stability threshold, the incomplete minimality search for Cases 2-5, and the fact that both latent size and entropy rise with input amplitude—are concerns about robustness and evidentiary support, not circularity. Under the hard rule requiring a specific reduction, no circular step can be exhibited.

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

No new physical entities are introduced. The paper's contribution relies on a set of hand-picked thresholds, architectures, and amplitudes, plus the domain assumption that the Schrödinger equation is the ground truth.

free parameters (4)
  • stability threshold = over 85% of R² above 0.85
    Used to define minimum stable architecture in Sec. III B; changing this threshold would change the reported minimal latent sizes and the complexity ranking.
  • architecture sizes = (5-5-3), (12-12-10-10), etc.
    Selected by hand from a few candidates; the minimum is only minimum among tested architectures, not a global optimum.
  • input amplitude values = A=1, 10, 1.5, 2.5 and A1=0.5, Am=1.7
    Case design; complexity ranking may be driven by amplitude since both latent size and permutation entropy increase with it.
  • permutation entropy embedding dimension = 4
    Used in Sec. III E via EntropyHub; the complexity measure depends on this choice.
assumptions (4)
  • domain assumption The Hamiltonian and time-dependent Schrödinger equation generate the ground-truth input-output mapping.
    QuTiP integration is assumed accurate and deterministic for N=10 chains.
  • domain assumption The input-output map h(t) to Y(t) is a deterministic causal function of the input time series.
    Justifies the TCN causal architecture; no noise or memory beyond the system state.
  • ad hoc to paper The latent dimension of a TCN autoencoder is a monotonic proxy for the intrinsic complexity of the mapping.
    Central claim in Sec. III C/E; not derived, only supported by five cases and comparison with permutation entropy.
  • domain assumption Amplitude-aware permutation entropy with embedding dimension 4 is a valid independent complexity measure.
    Taken from EntropyHub literature; used as ground truth for ranking in Sec. III E.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Input-Output Optics as a Causal Time Series Mapping: A Generative Machine Learning Solution." pith.science (2026). https://pith.science/paper/IYM2PRW7

@misc{pith2026241119897,
  author       = {Pith},
  title        = {Pith review of: Input-Output Optics as a Causal Time Series Mapping: A Generative Machine Learning Solution},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IYM2PRW7}},
  note         = {Machine review of arXiv:2411.19897}
}
read the original abstract

The response of many-body quantum systems to an optical pulse can be extremely challenging to model. Here we explore the use of neural networks, both traditional and generative, to learn and thus simulate the response of such a system from data. The quantum system can be viewed as performing a complex mapping from an input time-series (the optical pulse) to an output time-series (the systems response) which is often also an optical pulse. Using both the transverse and non-integrable Ising models as examples, we show that not only can temporal convolutional networks capture the input/output mapping generated by the system but can also be used to characterize the complexity of the mapping. This measure of complexity is provided by the size of the smallest latent space that is able to accurately model the mapping. We further find that a generative model, in particular a variational auto-encoder, significantly outperforms traditional auto-encoders at learning the complex response of many-body quantum systems. For the example that generated the most complex mapping, the variational auto-encoder produces outputs that have less than 10% error for more than 90% of inputs across our test data.

Figures

Figures reproduced from arXiv: 2411.19897 by the authors.

Figure 1
Figure 1. FIG. 1. Optics as Input-Output Problem: The input (red pulse) gets [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2. Schematic picture of time series modelling using [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. FIG. 3. Deep Neural Network Autoencoder Architechture [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figures from the paper (9 more)
Figure 4
Figure 4. Figure 4: FIG. 4. TCN Autoencoder Architecture [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: FIG. 5. Performance illustration of a temporal convolutional network [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]
Figure 7
Figure 7. Figure 7: FIG. 7. Performance illustration of a temporal convolutional net [PITH_FULL_IMAGE:figures/full_fig_p007_7.png]
Figure 9
Figure 9. Figure 9: FIG. 9. Performance illustration of a variational temporal convolu [PITH_FULL_IMAGE:figures/full_fig_p008_9.png]
Figure 10
Figure 10. Figure 10: FIG. 10. The mean normalized permutation probabilities (Pnorm) [PITH_FULL_IMAGE:figures/full_fig_p009_10.png]
Figure 11
Figure 11. Figure 11: Case (b): Architecture : (32−16−8−6), Number of pa￾rameters = 4,439. We implemented a CNN-based autoen￾coder model with a comparable number of parameters to the TCN-based model to ensure a fair comparison of perfor￾mance. The R 2 value shows slight improvement over Ca…
Figure 12
Figure 12. Figure 12: FIG. 12. Plot shows the [PITH_FULL_IMAGE:figures/full_fig_p010_12.png]
Figure 13
Figure 13. Figure 13: FIG. 13. This box plot visually summarizes the performance of dif [PITH_FULL_IMAGE:figures/full_fig_p011_13.png]
Figure 14
Figure 14. Figure 14: FIG. 14. This box plot compares the performance of three mod [PITH_FULL_IMAGE:figures/full_fig_p012_14.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

101 extracted references · 67 canonical work pages

  1. [1]

    S. H. Strogatz, Nonlinear Dynamics and Chaos (CRC Press, 2018)

    2018

  2. [2]

    is the model’s predicted output time series for the given input time series h(i). Similarly, for Case 5 TCN : Layer = 8, Filter = 3, Dilation = [2,4,8] TCN : Layer = 6, Filter = 3, Dilation = [2,4,8] Max Pooling : Size = 2 Max Pooling : Size = 2 Latent TCN : Layer = 6, Filter = 3,Dilation = [2,4,8] TCN : Layer = 6, Filter = 3, Dilation = [2,4,8] Up Sampli...

  3. [3]

    Amini, J

    K. Amini, J. Biegert, F. Calegari, A. Chac ´on, M. F. Ciap- pina, A. Dauphin, D. K. Efimov, C. Figueira De Moris- son Faria, K. Giergiel, P. Gniewek, A. S. Landsman, M. Le- siuk, M. Mandrysz, A. S. Maxwell, R. Moszy´nski, L. Ortmann, J. Antonio P´erez-Hern´andez, A. Pic ´on, E. Pisanty, J. Prauzner- Bechcicki, K. Sacha, N. Su ´arez, A. Za ¨ır, J. Zakrzews...

    2019

  4. [4]

    Remarkably, all mod- els achieved an R-squared value exceeding 0.95, with at least 88% of {R2 i |i ∈ TestSet} surpassing this threshold

    architecture with 3,221 parameters. Remarkably, all mod- els achieved an R-squared value exceeding 0.95, with at least 88% of {R2 i |i ∈ TestSet} surpassing this threshold. In all the cases treated, we ensured a similar train-test split of the data and kept the number of epochs the same. For all our training processes we used the swish activation function...

  5. [5]

    R. W. Boyd, The nonlinear optical susceptibility, in Nonlinear Optics (Elsevier, 2008) pp. 1–67

    2008

  6. [6]

    J. L. Krause, K. J. Schafer, and K. C. Kulander, High-order harmonic generation from atoms and ions in the high intensity regime, Physical Review Letters 68, 3535 (1992)

    1992

  7. [7]

    P. B. Corkum, Plasma perspective on strong field multiphoton ionization, Physical Review Letters 71, 1994 (1993)

    1993

  8. [8]

    Lewenstein, P

    M. Lewenstein, P. Balcou, M. Y . Ivanov, A. L’Huillier, and P. B. Corkum, Theory of high-harmonic generation by low- frequency laser fields, Physical Review A 49, 2117 (1994)

    1994

Show all 101 references
  1. [9]

    M. F. Ciappina, J. A. P ´erez-Hern´andez, A. S. Landsman, T. Zimmermann, M. Lewenstein, L. Roso, and F. Krausz, 13 Carrier-wave rabi-flopping signatures in high-order harmonic generation for alkali atoms, Physical Review Letters 114, 143902 (2015)

    2015

  2. [10]

    Krausz and M

    F. Krausz and M. Ivanov, Attosecond physics, Rev. Mod. Phys. 81, 163 (2009)

    2009

  3. [11]

    Vimal, M

    M. Vimal, M. Luttmann, T. Gadeyne, M. Guer, R. Cazali, D. Bresteau, F. Lepetit, O. Tcherbakoff, J.-F. Hergott, T. Au- guste, and T. Ruchon, Photon pathways and the nonperturba- tive scaling law of high harmonic generation, Physical Review Letters 131, 203402 (2023)

    2023

  4. [12]

    T. X. Luu, M. Garg, X. Yu, A. Liu, S. M. Teichmann, S. L. Cousin, H. J. W¨orner, and E. Goulielmakis, Extreme ultraviolet high-harmonic spectroscopy of solids, Nature 521, 498 (2015)

    2015

  5. [13]

    Ghimire, A

    S. Ghimire, A. D. DiChiara, E. Sistrunk, P. Agostini, L. F. Di- Mauro, and D. A. Reis, Observation of high-order harmonic generation in a bulk crystal, Nature Physics 7, 138 (2011)

    2011

  6. [14]

    Vampa, T

    G. Vampa, T. J. Hammond, N. Thir´e, B. E. Schmidt, F. L´egar´e, C. R. McDonald, T. Brabec, D. D. Klug, and P. B. Corkum, Linking high harmonics from gases and solids, Nature522, 462 (2015)

    2015

  7. [15]

    Vampa, C

    G. Vampa, C. R. McDonald, G. Orlando, D. D. Klug, P. B. Corkum, and T. Brabec, Theoretical analysis of high-harmonic generation in solids, Physical Review Letters 113, 073901 (2014)

    2014

  8. [16]

    Navarrete and U

    F. Navarrete and U. Thumm, Two-color-driven enhanced high- order harmonic generation in solids, Phys. Rev. A 102, 063123 (2020)

    2020

  9. [17]

    J. Li, J. Lu, A. Chew, S. Han, J. Li, Y . Wu, H. Wang, S. Ghimire, and Z. Chang, Attosecond science based on high harmonic generation from gases and solids, Nature Communications 11, 2748 (2020)

    2020

  10. [18]

    Ortmann, J

    L. Ortmann, J. A. P ´erez-Hern´andez, M. F. Ciappina, J. Sch ¨otz, A. Chac´on, G. Zeraouli, M. F. Kling, L. Roso, M. Lewenstein, and A. S. Landsman, Emergence of a higher energy structure in strong field ionization with inhomogeneous electric fields, Physical Review Letters 11...

    2017

  11. [19]

    Ghimire and D

    S. Ghimire and D. A. Reis, High-harmonic generation from solids, Nature Physics 15, 10 (2019)

    2019

  12. [20]

    Hohenleutner, F

    M. Hohenleutner, F. Langer, O. Schubert, M. Knorr, U. Huttner, S. W. Koch, M. Kira, and R. Huber, Real-time observation of in- terfering crystal electrons in high-harmonic generation, Nature 523, 572 (2015)

    2015

  13. [21]

    R. E. F. Silva, I. V . Blinov, A. N. Rubtsov, O. Smirnova, and M. Ivanov, High-harmonic spectroscopy of ultrafast many-body dynamics in strongly correlated systems, Nature Photonics 12, 266–270 (2018)

    2018

  14. [22]

    Vampa, B

    G. Vampa, B. Ghamsari, S. Siadat Mousavi, T. Hammond, A. Olivieri, E. Lisicka-Skrek, A. Y . Naumov, D. Villeneuve, A. Staudte, P. Berini, et al., Plasmon-enhanced high-harmonic generation from silicon, Nature Physics 13, 659 (2017)

    2017

  15. [23]

    Tancogne-Dejean, M

    N. Tancogne-Dejean, M. A. Sentef, and A. Rubio, Ultrafast modification of hubbard u in a strongly correlated material:ab initio high-harmonic generation in nio, Physical Review Letters 121, 10.1103/physrevlett.121.097402 (2018)

    2018 doi

  16. [24]

    Korobenko, S

    A. Korobenko, S. Saha, A. T. K. Godfrey, M. Gertsvolf, A. Y . Naumov, D. M. Villeneuve, A. Boltasseva, V . M. Shalaev, and P. B. Corkum, High-harmonic generation in metallic titanium nitride, Nature Communications 12, 4981 (2021)

    2021

  17. [25]

    Murakami, S

    Y . Murakami, S. Takayoshi, A. Koga, and P. Werner, High- harmonic generation in one-dimensional mott insulators, Phys- ical Review B 103, 10.1103/physrevb.103.035110 (2021)

    2021 doi

  18. [26]

    Orthodoxou, A

    C. Orthodoxou, A. Za ¨ır, and G. H. Booth, High harmonic gen- eration in two-dimensional mott insulators, npj Quantum Mate- rials 6, 10.1038/s41535-021-00377-8 (2021)

    2021 doi

  19. [27]

    R. K. Malla, D. Kim, D. E. Kim, A. Chac ´on, and W. J. M. Kort-Kamp, Ultrafast nonequilibrium dynamics and high- harmonic generation in two-dimensional quantum spin hall ma- terials, Physical Review Materials 7, 10.1103/physrevmateri- als.7.l051201 (2023)

    2023 doi

  20. [28]

    Takayoshi, Y

    S. Takayoshi, Y . Murakami, and P. Werner, High-harmonic gen- eration in quantum spin systems, Physical Review B99, 022410 (2019)

    2019

  21. [29]

    T. N. Ikeda and M. Sato, High-harmonic generation by electric polarization, spin current, and magnetization, Physical Review B 100, 10.1103/physrevb.100.214424 (2019)

    2019 doi

  22. [30]

    H. E. Stanley, Introduction to Phase Transitions and Critical Phenomena (Oxford University Press, 1971)

    1971

  23. [31]

    N. D. Klimkin, A. N. Rubtsov, and M. Ivanov, Coherent con- trol of symmetry breaking in transverse-field Ising chains using few-cycle pulses, Nanophotonics 10.1515/nanoph-2022-0811 (2023), publisher: De Gruyter

    2023 doi

  24. [32]

    Alcal `a, U

    J. Alcal `a, U. Bhattacharya, J. Biegert, M. Ciappina, U. Elu, T. Graß, P. T. Grochowski, M. Lewenstein, A. Palau, T. P. H. Sidiropoulos, T. Steinle, and I. Tyulnev, High-harmonic spec- troscopy of quantum phase transitions in a high-Tc supercon- ductor, Proceedings of the Nat...

    2022

  25. [33]

    Suzuki, Quantum Ising Phases and Transitions in Transverse Ising Models (Springer, 2012)

    S. Suzuki, Quantum Ising Phases and Transitions in Transverse Ising Models (Springer, 2012)

    2012

  26. [34]

    Sachdev, Quantum Phase Transitions(Cambridge University Press, 2011)

    S. Sachdev, Quantum Phase Transitions(Cambridge University Press, 2011)

    2011

  27. [35]

    B. K. Chakrabarti, A. Dutta, and P. Sen, Quantum Ising Phases and Transitions in Transverse Ising Models(Springer, 2008)

    2008

  28. [36]

    A. A. Ovchinnikov, D. V . Dmitriev, V . Y . Krivnov, and V . O. Cheranovskii, Antiferromagnetic ising chain in a mixed trans- verse and longitudinal magnetic field, Physical Review B 68, 10.1103/physrevb.68.214406 (2003)

    2003 doi

  29. [37]

    M. W. Johnson, M. H. S. Amin, S. Gildert, T. Lanting, F. Hamze, N. Dickson, et al., Quantum annealing with man- ufactured spins, Nature 473, 194 (2011)

    2011

  30. [38]

    Dziarmaga, Dynamics of a quantum phase transition: Exact solution of the quantum ising model, Physical Review Letters 95, 10.1103/physrevlett.95.245701 (2005)

    J. Dziarmaga, Dynamics of a quantum phase transition: Exact solution of the quantum ising model, Physical Review Letters 95, 10.1103/physrevlett.95.245701 (2005)

    2005 doi

  31. [39]

    S. Bai, J. Z. Kolter, and V . Koltun, An empirical evaluation of generic convolutional and recurrent networks for sequence modeling, CoRR abs/1803.01271 (2018), 1803.01271

    2018 arXiv

  32. [40]

    Pandey and D

    A. Pandey and D. Wang, Tcnn: Temporal convolutional neu- ral network for real-time speech enhancement in the time do- main, in ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) (IEEE, 2019)

    2019

  33. [41]

    C. Lea, M. D. Flynn, R. Vidal, A. Reiter, and G. D. Hager, Temporal convolutional networks for action segmentation and detection, Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition , 156 (2017)

    2017

  34. [42]

    Y . Li, L. Xiao, H. Wei, D. Li, and X. Li, A comparative study of lstm and temporal convolutional network models for semisub- mersible platform wave runup prediction, Journal of Offshore Mechanics and Arctic Engineering 147, 10.1115/1.4063266 (2024)

    2024 doi

  35. [43]

    van den Oord, S

    A. van den Oord, S. Dieleman, H. Zen, K. Simonyan, O. Vinyals, A. Graves, N. Kalchbrenner, A. Senior, and K. Kavukcuoglu, Wavenet: A generative model for raw audio, arXiv preprint arXiv:1609.03499 (2016)

    2016 arXiv

  36. [44]

    D. Cer, Y . Yang, S.-y. Kong, N. Hua, N. Limtiaco, R. St. John, N. Constant, M. Guajardo-C ´espedes, S. Yuan, C. Tar, et al. , Tcn: Temporal convolutional networks for the universal sen- tence encoder, in Proceedings of the 56th Annual Meeting of the Association for Computatio...

    2018

  37. [45]

    Yan, S.-S

    J.-Z. Yan, S.-S. Zhao, W.-D. Lan, S.-Y . Li, S.-S. Zhou, J.-G. Chen, J.-Y . Zhang, and Y .-J. Yang, Calculation of high-order harmonic generation of atoms and molecules by combining time series prediction and neural networks, Optics Express 30, 35444 (2022)

    2022

  38. [46]

    J. Yan, L. Mu, L. Wang, R. Ranjan, and A. Y . Zomaya, Tempo- ral convolutional networks for the advance prediction of enso, Scientific Reports 10, 10.1038/s41598-020-65070-5 (2020). 14

    2020 doi

  39. [47]

    Hochreiter and J

    S. Hochreiter and J. Schmidhuber, Long short-term memory, Neural computation 9, 1735 (1997)

    1997

  40. [48]

    Gross, H

    P. Gross, H. Singh, H. Rabitz, K. Mease, and G. M. Huang, Inverse quantum-mechanical control: A means for design and a test of intuition, Physical Review A 47, 4593 (1993)

    1993

  41. [49]

    C. K. Ong, G. M. Huang, T. J. Tarn, and J. W. Clark, Invert- ibility of quantum-mechanical control systems, Mathematical Systems Theory 17, 335 (1984)

    1984

  42. [50]

    J. W. Clark, C. K. Ong, T. J. Tarn, and G. M. Huang, Quan- tum nondemolition filters, Mathematical Systems Theory 18, 33 (1985)

    1985

  43. [51]

    Rothman, T.-S

    A. Rothman, T.-S. Ho, and H. Rabitz, Observable-preserving control of quantum dynamics over a family of related systems, Physical Review A 72, 023416 (2005)

    2005

  44. [52]

    W. Zhu, M. Smit, and H. Rabitz, Managing singular behavior in the tracking control of quantum dynamical observables, The Journal of Chemical Physics 110, 1905 (1999)

    1999

  45. [53]

    Magann, T.-S

    A. Magann, T.-S. Ho, and H. Rabitz, Singularity-free quantum tracking control of molecular rotor orientation, Physical Review A 98, 043429 (2018)

    2018

  46. [54]

    McCaul, C

    G. McCaul, C. Orthodoxou, K. Jacobs, G. H. Booth, and D. I. Bondar, Driven Imposters: Controlling Expectations in Many- Body Systems, Physical Review Letters 124, 183201 (2020)

    2020

  47. [55]

    A. B. Magann, T.-S. Ho, C. Arenz, and H. A. Rabitz, Quantum tracking control of the orientation of symmetric-top molecules, Physical Review A 108, 033106 (2023)

    2023

  48. [56]

    A. G. Campos, D. I. Bondar, R. Cabrera, and H. A. Rab- itz, How to Make Distinct Dynamical Systems Appear Spec- trally Identical, Physical Review Letters 118, 10.1103/Phys- RevLett.118.083201 (2017)

    2017 doi

  49. [57]

    Y . Shen, E. Liao, T. Tai, Y . Liao, C. Sun, C. Lee, S. See, and H. Chen, Deep learning-based photodamage reduction on har- monic generation microscope at low-level optical power, Jour- nal of Biophotonics 17, 10.1002/jbio.202300285 (2023)

    2023 doi

  50. [58]

    McCaul, A

    G. McCaul, A. F. King, and D. I. Bondar, Optical Indistin- guishability via Twinning Fields, Physical Review Letters 127, 113201 (2021)

    2021

  51. [59]

    Lohani, E

    S. Lohani, E. M. Knutson, W. Zhang, and R. T. Glasser, Disper- sion characterization and pulse prediction with machine learn- ing, OSA Continuum 2, 3438 (2019)

    2019

  52. [60]

    Lytova, M

    M. Lytova, M. Spanner, and I. Tamblyn, Deep learning and high harmonic generation, Canadian Journal of Physics 101, 132–140 (2023)

    2023

  53. [61]

    N. D. Klimkin, A. Jim´enez-Gal´an, R. E. F. Silva, and M. Ivanov, Symmetry-aware deep neural networks for high harmonic spec- troscopy in solids, Optics Express 31, 20559 (2023)

    2023

  54. [62]

    Mihailescu, A new approach to theoretical investigations of high harmonics generation by means of fs laser interaction with overdense plasma layers

    A. Mihailescu, A new approach to theoretical investigations of high harmonics generation by means of fs laser interaction with overdense plasma layers. combining particle-in-cell sim- ulations with machine learning., Journal of Instrumentation 11 (12), C12004–C12004

  55. [63]

    Brunner, A

    C. Brunner, A. Duensing, C. Schr ¨oder, M. Mittermair, V . Golkov, M. Pollanka, D. Cremers, and R. Kienberger, Deep learning in attosecond metrology, Optics Express 30, 15669 (2022)

    2022

  56. [64]

    J. M. Pablos-Mar´ın, J. Serrano, and C. Hern´andez-Garc´ıa, Sim- ulating macroscopic high-order harmonic generation driven by structured laser beams using artificial intelligence, Computer Physics Communications 291, 108823 (2023)

    2023

  57. [65]

    Zahavy, A

    T. Zahavy, A. Dikopoltsev, D. Moss, G. I. Haham, O. Cohen, S. Mannor, and M. Segev, Deep learning reconstruction of ul- trashort pulses, Optica 5, 666 (2018)

    2018

  58. [66]

    Mitra and H

    A. Mitra and H. Rabitz, Identifying mechanisms in the control of quantum dynamics through Hamiltonian encoding, Physical Review A 67, 033407 (2003)

    2003

  59. [67]

    Sanchez-Gonzalez, P

    A. Sanchez-Gonzalez, P. Micaelli, C. Olivier, T. R. Barillot, M. Ilchen, A. A. Lutman, A. Marinelli, T. Maxwell, A. Achner, M. Ag˚aker, N. Berrah, C. Bostedt, J. D. Bozek, J. Buck, P. H. Bucksbaum, S. C. Montero, B. Cooper, J. P. Cryan, M. Dong, R. Feifel, L. J. Frasinski, H. ...

    2017

  60. [68]

    Kumar Giri, U

    S. Kumar Giri, U. Saalmann, and J. M. Rost, Purifying Electron Spectra from Noisy Pulses with Machine Learning Using Syn- thetic Hamilton Matrices, Physical Review Letters124, 113201 (2020)

    2020

  61. [69]

    Azami and J

    H. Azami and J. Escudero, Amplitude-aware permutation en- tropy: Illustration in spike detection and signal segmentation, Computer Methods and Programs in Biomedicine 128, 40–51 (2016)

    2016

  62. [70]

    Rey-de Castro, R

    R. Rey-de Castro, R. Cabrera, D. I. Bondar, and H. Ra- bitz, Time-resolved quantum process tomography using Hamiltonian-encoding and observable-decoding, New Journal of Physics 15, 025032 (2013)

    2013

  63. [71]

    Mal´y, J

    P. Mal´y, J. L¨uttig, P. A. Rose, A. Turkin, C. Lambert, J. J. Krich, and T. Brixner, Separating single- from multi-particle dynamics in nonlinear spectroscopy, Nature 616, 280 (2023)

    2023

  64. [72]

    H. Kim, T. N. Ikeda, and D. A. Huse, Testing whether all eigen- states obey the eigenstate thermalization hypothesis, Physical Review E 90, 052105 (2014)

    2014

  65. [73]

    Mukai, H

    Y . Mukai, H. Hirori, T. Yamamoto, H. Kageyama, and K. Tanaka, Antiferromagnetic resonance excitation by terahertz magnetic field resonantly enhanced with split ring resonator, Applied Physics Letters 105, 10.1063/1.4890475 (2014)

    2014 doi

  66. [74]

    Chang and N

    Z. Chang and N. Wu, Dynamics of entanglement in the trans- verse ising model, Physical Review A 81, 022312 (2010)

    2010

  67. [75]

    Hughes, On the mean accuracy of statistical pattern recog- nizers, IEEE Transactions on Information Theory 14, 55–63 (1968)

    G. Hughes, On the mean accuracy of statistical pattern recog- nizers, IEEE Transactions on Information Theory 14, 55–63 (1968)

    1968

  68. [76]

    Johansson, P

    J. Johansson, P. Nation, and F. Nori, Qutip 2: A python frame- work for the dynamics of open quantum systems, Computer Physics Communications 184, 1234–1240 (2013)

    2013

  69. [77]

    Johansson, P

    J. Johansson, P. Nation, and F. Nori, Qutip: An open-source python framework for the dynamics of open quantum systems, Computer Physics Communications 183, 1760–1772 (2012)

    2012

  70. [78]

    Input-output optics as a causal time-series mapping: A genera- tive machine learning solution, GitHub Link (2025)

    2025

  71. [79]

    Wang and V

    Q. Wang and V . Megalooikonomou, A dimensionality reduction technique for efficient time series similarity analysis, Informa- tion Systems 33, 115–132 (2008)

    2008

  72. [80]

    Chiarot and C

    G. Chiarot and C. Silvestri, Time series compression survey, ACM Computing Surveys 55, 1–32 (2023)

    2023

  73. [81]

    Chalapathy and S

    R. Chalapathy and S. Chawla, Deep learning for anomaly de- tection: A survey, arXiv preprint arXiv:1901.03407 (2019)

    2019 arXiv

  74. [82]

    D. P. Kingma and M. Welling, Auto-encoding variational bayes (2013), arXiv:1312.6114

    2013 arXiv

  75. [83]

    Strub, J

    F. Strub, J. Mary, and R. Gaudel, Deep autoencoders for collab- orative filtering, arXiv preprint arXiv:1606.07792 (2016). 15

    2016 arXiv

  76. [84]

    P. J. Huber, Robust estimation of a location parameter, The An- nals of Mathematical Statistics 35, 73 (1964)

    1964

  77. [85]

    H. V . Pham, S. Qian, J. Wang, T. Lutellier, J. Rosenthal, L. Tan, Y . Yu, and N. Nagappan, Problems and opportunities in train- ing deep learning software systems: An analysis of variance, in 2020 35th IEEE/ACM International Conference on Automated Software Engineering (ASE) ...

    2020

  78. [86]

    Ramachandran, B

    P. Ramachandran, B. Zoph, and Q. V . Le, Searching for activa- tion functions (2017), arXiv:1710.05941

    2017 arXiv

  79. [87]

    Bandt and B

    C. Bandt and B. Pompe, Permutation entropy: a natural com- plexity measure for time series, Physical Review Letters 88, 174102 (2002)

    2002

  80. [88]

    S. L. Smith, P.-J. Kindermans, C. Ying, and Q. V . Le, Don’t decay the learning rate, increase the batch size (2017), arXiv:1711.00489

    2017 arXiv

  81. [89]

    D. P. Kingma, M. Welling, et al., An introduction to variational autoencoders, Foundations and Trends in Machine Learning12, 307 (2019)

    2019

  82. [90]

    Pincus, Approximate entropy as a measure of system com- plexity, Proc

    S. Pincus, Approximate entropy as a measure of system com- plexity, Proc. Natl. Acad. Sci. U. S. A. 88, 2297 (1991)

    1991

  83. [91]

    J. M. Amig ´o, M. B. Kennel, and L. Kocarev, The permutation entropy rate equals the metric entropy rate for ergodic informa- tion sources and ergodic dynamical systems, Physical Review Letters 94, 144102 (2005)

    2005

  84. [92]

    M. W. Flood and B. Grimm, Entropyhub: An open-source toolkit for entropic time series analysis, PLOS ONE 16, e0259448 (2021)

    2021

  85. [93]

    Rosso, S

    O. Rosso, S. Blanco, J. Yordanova, V . Kolev, A. Figliola, and M. e. a. Sch ¨urmann, Wavelet entropy: a new tool for analysis of short duration brain electrical signals, J. Neurosci. Methods 105, 65 (2001)

    2001

  86. [94]

    Richman and J

    J. Richman and J. Moorman, Physiological time-series analysis using approximate entropy and sample entropy, Am. J. Physiol. Heart Circ. Physiol. 278, H2039 (2000)

    2000

  87. [95]

    W. Chen, Z. Wang, H. Xie, and W. Yu, Characterization of sur- face emg signal based on fuzzy entropy, IEEE Trans. Neural Syst. Rehabil. Eng. 15, 266 (2007)

    2007

  88. [96]

    Zanin, L

    M. Zanin, L. Zunino, O. Rosso, and D. Papo, Permutation en- tropy and its main biomedical and econophysics applications: a review, Entropy14, 1553 (2012)

    2012

  89. [97]

    Holzinger, M

    A. Holzinger, M. H ¨ortenhuber, C. Mayer, M. Bachler, S. Wassertheurer, and A. e. a. Pinho, On entropy-based data mining, in Interactive Knowledge Discovery and Data Mining in Biomedical Informatics , V ol. 8401, edited by A. Holzinger and I. Jurisica (Springer, Berlin, Heidel...

    2014

  90. [98]

    Keller, A

    K. Keller, A. Unakafov, and V . Unakafova, Ordinal patterns, entropy, and eeg, Entropy 16, 6212 (2014)

    2014

  91. [100]

    Y .-J. Yuan, X. Wang, Z.-T. Huang, and Z.-C. Sha, Detection of radio transient signal based on permutation entropy and glrt, Wireless Personal Communications 82, 1047 (2015)

    2015

  92. [101]

    G. Zhu, Y . Li, P. Wen, and S. Wang, Classifying epileptic eeg signals with delay permutation entropy and multi-scale k- means, in Signal and Image Analysis for Biomedical and Life Sciences, V ol. 823, edited by C. Sun, T. Bednarz, T. Pham, P. Vallotton, and D. Wang (Springer ...

  93. [512]

    ,yi 512)

    with the correspond- ing output vector (yi 1,yi 2, . . . ,yi 512). After splitting the dataset into training n h(i) → Y(i)|i ∈ TrainSet o and testingn h(i) → Y(i)|i ∈ TestSet o samples, we train a TCN model on the training sample. The obtained model can be used to produce the ...

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.