pith. sign in

arxiv: 2408.00222 · v2 · pith:MUUPU7L5new · submitted 2024-08-01 · ⚛️ physics.chem-ph

Non-markovian neural quantum propagator and its application to the simulation of ultrafast nonlinear spectra

Jiaji Zhang , Lipeng Chen This is my paper

Pith reviewed 2026-05-23 22:42 UTC · model grok-4.3

classification ⚛️ physics.chem-ph
keywords neural networkquantum propagatorhierarchical equations of motionopen quantum systemsFenna-Matthews-Olson complexnon-Markovian dynamicsultrafast spectra
0
0 comments X

The pith

A neural network can evolve quantum states under non-Markovian rules for arbitrarily long times without repeated iterations.

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

The paper develops a machine-learning model to solve the hierarchical equations of motion for dissipative quantum dynamics. The model uses a neural network to serve as a propagator that evolves any initial quantum state over long durations. It is applied to population dynamics and linear and two-dimensional spectra of the Fenna-Matthews-Olson complex. If correct, this removes the computational cost of iterative solutions at each time step while preserving non-Markovian and nonperturbative effects.

Core claim

The authors propose a neural quantum propagator model based on neural network architecture that serves as a universal solver for the hierarchical equations of motion. It avoids time-consuming iterations and evolves any initial quantum state for arbitrarily long times, demonstrated by simulating population dynamics and spectra in the Fenna-Matthews-Olson complex.

What carries the argument

The neural quantum propagator, a neural network trained on hierarchical equations of motion data to replace iterative propagation steps.

If this is right

  • Population dynamics of open quantum systems become computable without iterative calculations at each time step.
  • Linear and two-dimensional spectra of molecular complexes can be obtained for longer evolution periods.
  • Non-Markovian effects remain captured in a numerically exact manner through the trained network.
  • The same model applies to any initial quantum state without retraining for each new starting condition.

Where Pith is reading between the lines

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

  • The approach could reduce the cost of exploring parameter spaces in photosynthetic or chemical systems.
  • Training data from multiple environments might allow the network to handle varied system-bath couplings.
  • Extension to real-time experimental feedback loops could become practical if inference speed is high enough.

Load-bearing premise

A neural network trained on hierarchical equations of motion data will generalize accurately to systems such as the Fenna-Matthews-Olson complex and maintain fidelity over long evolution times.

What would settle it

Direct numerical comparison of population dynamics and spectra produced by the neural propagator against standard hierarchical equations of motion results for the Fenna-Matthews-Olson complex at long times.

Figures

Figures reproduced from arXiv: 2408.00222 by Jiaji Zhang, Lipeng Chen.

Figure 1
Figure 1. Figure 1: The architecture of (a) the NQP model, and (b) [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 3
Figure 3. Figure 3: Population dynamics computed using NQP model [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: The linear spectrum R (1)(ω) evaluated from the NQP model (blue line) and the RK4 method (red line), re￾spectively. C. Linear spectra Next, we apply our NQP model to simulate the linear and third-order response functions as defined in Eqs. (10) and (12). We choose the transition dipole operator as µˆ = X 7 j=1 µj (|j⟩⟨g| + |g⟩⟨j|), (21) where µj is the transition dipole moment of j-th pig￾ment. The system … view at source ↗
Figure 6
Figure 6. Figure 6: The rephasing (a, c) and non-rephasing (b, d) parts [PITH_FULL_IMAGE:figures/full_fig_p006_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: The rephasing (a, c) and non-rephasing (b, d) [PITH_FULL_IMAGE:figures/full_fig_p006_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: The rephasing (a, c) and non-rephasing (b, d) [PITH_FULL_IMAGE:figures/full_fig_p007_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: The relative error between the NQP model and the [PITH_FULL_IMAGE:figures/full_fig_p007_9.png] view at source ↗
read the original abstract

The accurate solution of dissipative quantum dynamics plays an important role on the simulation of open quantum systems. Here we propose a machine-learning-based universal solver for the hierarchical equations of motion, one of the most widely used approaches which takes into account non-markovian effects and nonperturbative system-environment interactions in a numerically exact manner. We develop a neural quantum propagator model by utilizing the neural network architecture, which avoids time-consuming iterations and can be used to evolve any initial quantum state for arbitrarily long times. To demonstrate the efficacy of our model, we apply it to the simulation of population dynamics and linear and two-dimensional spectra of the Fenna-Matthews-Olson complex.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

3 major / 2 minor

Summary. The manuscript proposes a neural-network-based quantum propagator trained to reproduce solutions of the hierarchical equations of motion (HEOM) for non-Markovian open quantum systems. It claims this model functions as a universal solver that evolves arbitrary initial density matrices for arbitrarily long times without iterative computations, and demonstrates its use for population dynamics and linear/two-dimensional spectra of the Fenna-Matthews-Olson complex.

Significance. If the generalization and stability claims hold with quantitative validation, the approach could provide a fast surrogate for HEOM in long-time simulations of complex open quantum systems such as photosynthetic complexes, potentially enabling broader exploration of nonlinear spectra. The work correctly identifies the computational bottleneck of iterative HEOM and attempts to address it via machine learning, but the absence of reported metrics leaves the practical gain unclear.

major comments (3)
  1. [Abstract / Application to FMO] Abstract and application section: the claim that the model 'avoids time-consuming iterations' and can 'evolve any initial quantum state for arbitrarily long times' is presented without any quantitative metrics (e.g., population errors, spectral fidelity, wall-time comparisons), error bars, or direct comparisons to HEOM or other propagators, rendering the efficacy assertion impossible to evaluate.
  2. [Model development and results] The neural propagator is trained directly on HEOM trajectories for the FMO complex; consequently the 'universal solver' reduces to a learned approximation of the training method rather than an independent derivation. No out-of-distribution tests (different initial coherences, altered bath parameters, or evolution times substantially exceeding the training window) are described, which directly undermines the central claim of stable propagation for arbitrary states and times.
  3. [Generalization discussion] The weakest assumption—that the learned map remains contractive and free of accumulating errors beyond the HEOM training distribution—is not tested; the manuscript therefore provides no evidence that the network can serve as a reliable long-time propagator independent of the original HEOM solver.
minor comments (2)
  1. [Methods] Notation for the neural-network architecture and loss function should be defined explicitly with equations rather than left implicit.
  2. [Figures] Figure captions for spectra should include the precise HEOM truncation level and time-step used for the reference data.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for the constructive feedback on our manuscript. We address each major comment below and will revise the manuscript to strengthen the presentation of results and validation where appropriate.

read point-by-point responses

  1. Referee: [Abstract / Application to FMO] Abstract and application section: the claim that the model 'avoids time-consuming iterations' and can 'evolve any initial quantum state for arbitrarily long times' is presented without any quantitative metrics (e.g., population errors, spectral fidelity, wall-time comparisons), error bars, or direct comparisons to HEOM or other propagators, rendering the efficacy assertion impossible to evaluate.

    Authors: We agree that quantitative support is essential for evaluating the claims. The revised manuscript will include explicit metrics such as mean absolute errors in populations, spectral fidelity measures (e.g., overlap or RMSE for linear and 2D spectra), error bars from multiple runs, and wall-time comparisons against direct HEOM propagation for the FMO simulations. revision: yes

  2. Referee: [Model development and results] The neural propagator is trained directly on HEOM trajectories for the FMO complex; consequently the 'universal solver' reduces to a learned approximation of the training method rather than an independent derivation. No out-of-distribution tests (different initial coherences, altered bath parameters, or evolution times substantially exceeding the training window) are described, which directly undermines the central claim of stable propagation for arbitrary states and times.

    Authors: The model is trained on HEOM data for FMO as the target application, which is standard for developing a surrogate solver. The architecture itself is not system-specific and learns a general time-evolution map from the HEOM solutions. We acknowledge the value of out-of-distribution tests and will add validation on varied initial coherences and bath parameters in the revision. For times beyond the training window, the repeated application is the intended use case; we will include additional long-time stability checks against HEOM where computationally feasible. revision: partial

  3. Referee: [Generalization discussion] The weakest assumption—that the learned map remains contractive and free of accumulating errors beyond the HEOM training distribution—is not tested; the manuscript therefore provides no evidence that the network can serve as a reliable long-time propagator independent of the original HEOM solver.

    Authors: We agree that explicit testing of error accumulation and contractivity is important for supporting long-time use. In the revision we will add analysis of the learned map's properties (e.g., norm preservation or error growth over extended propagations) and direct comparisons to HEOM for trajectories longer than the training data to address this point. revision: yes

Circularity Check

0 steps flagged

No significant circularity; NN is explicit surrogate model for HEOM

full rationale

The paper proposes training a neural network on trajectories generated by the hierarchical equations of motion (HEOM) to create a fast propagator for open quantum systems. This is a standard data-driven approximation technique whose outputs are, by design, interpolations or extrapolations of the training data; the abstract and description make no claim that the NN derives HEOM results from first principles or that any 'prediction' is independent of the fitted HEOM input. No self-definitional equations, fitted-input-called-prediction steps, or load-bearing self-citations appear. Generalization performance is a separate empirical question, not a circularity issue.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

The central claim rests on the HEOM framework as ground truth and on the ability of a neural network to learn its propagator; the neural network weights are free parameters fitted to HEOM-generated data.

free parameters (1)
  • Neural network weights and architecture hyperparameters
    Parameters fitted during training to approximate the HEOM time-evolution operator.
axioms (1)
  • domain assumption The hierarchical equations of motion provide a numerically exact description of the non-Markovian dissipative dynamics.
    The paper takes HEOM as the reference method whose solutions the neural model must reproduce.

pith-pipeline@v0.9.0 · 5639 in / 1285 out tokens · 23975 ms · 2026-05-23T22:42:01.407570+00:00 · methodology

discussion (0)

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

Reference graph

Works this paper leans on

53 extracted references · 53 canonical work pages · 2 internal anchors

  1. [1]

    author author M. F. \ Gelin , author L. Chen ,\ and\ author W. Domcke ,\ https://doi.org/10.1021/acs.chemrev.2c00329 journal journal Chem. Rev. \ volume 122 ,\ pages 17339--17396 ( year 2022 ) NoStop

    work page doi:10.1021/acs.chemrev.2c00329 2022

  2. [2]

    & Martin, F

    author author M. Nisoli , author P. Decleva , author F. Calegari , author A. Palacios ,\ and\ author F. Martín ,\ https://doi.org/10.1021/acs.chemrev.6b00453 journal journal Chem. Rev. \ volume 117 ,\ pages 10760--10825 ( year 2017 ) NoStop

    work page doi:10.1021/acs.chemrev.6b00453 2017

  3. [3]

    Mukamel ,\ https://doi.org/10.1146/annurev.physchem.51.1.691 journal journal Annu

    author author S. Mukamel ,\ https://doi.org/10.1146/annurev.physchem.51.1.691 journal journal Annu. Rev. Phys. Chem. \ volume 51 ,\ pages 691–729 ( year 2000 ) NoStop

    work page doi:10.1146/annurev.physchem.51.1.691 2000

  4. [4]

    Maiuri , author M

    author author M. Maiuri , author M. Garavelli ,\ and\ author G. Cerullo ,\ https://doi.org/10.1021/jacs.9b10533 journal journal J. Am. Chem. Soc. \ volume 142 ,\ pages 3–15 ( year 2019 ) NoStop

    work page doi:10.1021/jacs.9b10533 2019

  5. [5]

    author author K. E. \ Dorfman , author F. Schlawin ,\ and\ author S. Mukamel ,\ https://doi.org/10.1103/RevModPhys.88.045008 journal journal Rev. Mod. Phys. \ volume 88 ,\ pages 045008 ( year 2016 ) NoStop

    work page doi:10.1103/revmodphys.88.045008 2016

  6. [6]

    Fresch , author F

    author author E. Fresch , author F. V. A. \ Camargo , author Q. Shen , author C. C. \ Bellora , author T. Pullerits , author G. S. \ Engel , author G. Cerullo ,\ and\ author E. Collini ,\ @noop journal journal Nat. Rev. Methods Primers \ volume 3 ( year 2023 ) NoStop

    work page 2023

  7. [7]

    author author T. A. A. \ Oliver ,\ https://doi.org/10.1098/rsos.171425 journal journal R. Soc. Open Sci. \ volume 5 ,\ pages 171425 ( year 2018 ) NoStop

    work page doi:10.1098/rsos.171425 2018

  8. [8]

    author author G. S. \ Schlau-Cohen , author A. Ishizaki ,\ and\ author G. R. \ Fleming ,\ https://doi.org/https://doi.org/10.1016/j.chemphys.2011.04.025 journal journal Chem. Phys. \ volume 386 ,\ pages 1--22 ( year 2011 ) NoStop

    work page doi:10.1016/j.chemphys.2011.04.025 2011

  9. [9]

    author author N. S. \ Ginsberg , author Y.-C. \ Cheng ,\ and\ author G. R. \ Fleming ,\ https://doi.org/10.1021/ar9001075 journal journal Acc. Chem. Res. \ volume 42 ,\ pages 1352–1363 ( year 2009 ) NoStop

    work page doi:10.1021/ar9001075 2009

  10. [10]

    author author G. D. \ Scholes , author G. R. \ Fleming , author A. Olaya-Castro ,\ and\ author R. van Grondelle ,\ https://doi.org/10.1038/nchem.1145 journal journal Nat. Chem. \ volume 3 ,\ pages 763–774 ( year 2011 ) NoStop

    work page doi:10.1038/nchem.1145 2011

  11. [11]

    Kullmann , author S

    author author M. Kullmann , author S. Ruetzel , author J. Buback , author P. Nuernberger ,\ and\ author T. Brixner ,\ https://doi.org/10.1021/ja2032037 journal journal J. Am. Chem. Soc. \ volume 133 ,\ pages 13074--13080 ( year 2011 ) NoStop

    work page doi:10.1021/ja2032037 2011

  12. [12]

    author author E. A. \ Arsenault , author P. Bhattacharyya , author Y. Yoneda ,\ and\ author G. R. \ Fleming ,\ @noop journal journal J. Chem. Phys. \ volume 155 ( year 2021 ) NoStop

    work page 2021

  13. [13]

    Kim , author J

    author author J. Kim , author J. Jeon , author T. H. \ Yoon ,\ and\ author M. Cho ,\ @noop journal journal Nat. Comm. \ volume 11 ( year 2020 ) NoStop

    work page 2020

  14. [14]

    Ruetzel , author M

    author author S. Ruetzel , author M. Kullmann , author J. Buback , author P. Nuernberger ,\ and\ author T. Brixner ,\ @noop journal journal Phys. Rev. Lett \ volume 110 ( year 2013 ) NoStop

    work page 2013

  15. [15]

    Cho ,\ https://doi.org/10.1007/978-981-13-9753-0 title Coherent Multidimensional Spectroscopy \ ( publisher Springer Singapore ,\ year 2019 ) NoStop

    author author M. Cho ,\ https://doi.org/10.1007/978-981-13-9753-0 title Coherent Multidimensional Spectroscopy \ ( publisher Springer Singapore ,\ year 2019 ) NoStop

    work page doi:10.1007/978-981-13-9753-0 2019

  16. [16]

    Mukamel ,\ @noop title Principles of Nonlinear Optical Spectroscopy \ ( publisher Oxford University Press ,\ year 1995 ) NoStop

    author author S. Mukamel ,\ @noop title Principles of Nonlinear Optical Spectroscopy \ ( publisher Oxford University Press ,\ year 1995 ) NoStop

    work page 1995

  17. [17]

    \ Breuer \ and\ author F

    author author H.-P. \ Breuer \ and\ author F. Petruccione ,\ @noop title The Theory of Open Quantum Systems \ ( publisher Oxford University Press ,\ year 2007 ) NoStop

    work page 2007

  18. [18]

    Weiss ,\ @noop title Quantum Dissipative Systems ,\ edition 4th \ ed.\ ( publisher World Scientific ,\ year 2012 ) NoStop

    author author U. Weiss ,\ @noop title Quantum Dissipative Systems ,\ edition 4th \ ed.\ ( publisher World Scientific ,\ year 2012 ) NoStop

    work page 2012

  19. [19]

    Tanimura, Numerically “exact” approach to open quantum dynamics: The hierarchical equations of mo- tion (heom), The Journal of Chemical Physics153, 10.1063/5.0011599 (2020)

    author author Y. Tanimura ,\ https://doi.org/10.1063/5.0011599 journal journal J. Chem. Phys. \ volume 153 ,\ pages 020901 ( year 2020 ) NoStop

    work page doi:10.1063/5.0011599 2020

  20. [20]

    Ye , author X

    author author L. Ye , author X. Wang , author D. Hou , author R. Xu , author X. Zheng ,\ and\ author Y. Yan ,\ https://doi.org/10.1002/wcms.1269 journal journal WIREs Comput. Mol. Sci. \ volume 6 ,\ pages 608–638 ( year 2016 ) NoStop

    work page doi:10.1002/wcms.1269 2016

  21. [21]

    Zhang \ and\ author Y

    author author J. Zhang \ and\ author Y. Tanimura ,\ @noop journal journal J. Chem. Phys. \ volume 156 ( year 2022 ) NoStop

    work page 2022

  22. [22]

    author author P. E. \ Kloeden \ and\ author E. Platen ,\ https://doi.org/10.1007/978-3-662-12616-5 title Numerical Solution of Stochastic Differential Equations \ ( publisher Springer Berlin Heidelberg ,\ year 1992 ) NoStop

    work page doi:10.1007/978-3-662-12616-5 1992

  23. [23]

    Yan , author M

    author author Y. Yan , author M. Xu , author T. Li ,\ and\ author Q. Shi ,\ https://doi.org/10.1063/5.0050720 journal journal J. Chem. Phys. \ volume 154 ,\ pages 194104 ( year 2021 ) NoStop

    work page doi:10.1063/5.0050720 2021

  24. [24]

    Ke ,\ @noop journal journal J

    author author Y. Ke ,\ @noop journal journal J. Chem. Phys. \ volume 158 ( year 2023 ) NoStop

    work page 2023

  25. [25]

    Kimura \ and\ author Y

    author author A. Kimura \ and\ author Y. Fujihashi ,\ https://doi.org/10.1063/1.4901431 journal journal J. Chem. Phys. \ volume 141 ,\ pages 194110 ( year 2014 ) NoStop

    work page doi:10.1063/1.4901431 2014

  26. [26]

    author author A. W. \ Schlimgen , author K. Head-Marsden , author L. M. \ Sager , author P. Narang ,\ and\ author D. A. \ Mazziotti ,\ https://doi.org/10.1103/PhysRevLett.127.270503 journal journal Phys. Rev. Lett \ volume 127 ,\ pages 270503 ( year 2021 ) NoStop

    work page doi:10.1103/physrevlett.127.270503 2021

  27. [27]

    Liu , author Z.-H

    author author W. Liu , author Z.-H. \ Chen , author Y. Su , author Y. Wang ,\ and\ author W. Dou ,\ https://doi.org/10.1063/5.0170512 journal journal J. Chem. Phys. \ volume 159 ,\ pages 144110 ( year 2023 ) NoStop

    work page doi:10.1063/5.0170512 2023

  28. [28]

    LeCun, Y

    author author Y. LeCun , author Y. Bengio ,\ and\ author G. Hinton ,\ https://doi.org/10.1038/nature14539 journal journal Nature \ volume 521 ,\ pages 436--444 ( year 2015 ) NoStop

    work page doi:10.1038/nature14539 2015

  29. [29]

    Hermann , author J

    author author J. Hermann , author J. Spencer , author K. Choo , author A. Mezzacapo , author W. M. C. \ Foulkes , author D. Pfau , author G. Carleo ,\ and\ author F. Noé ,\ https://doi.org/10.1038/s41570-023-00516-8 journal journal Nat. Rev. Chem. \ volume 7 ,\ pages 692–709 ( year 2023 ) NoStop

    work page doi:10.1038/s41570-023-00516-8 2023

  30. [30]

    Learning nonlinear operators via DeepONet based on the universal approximation theorem of operators

    author author L. Lu , author P. Jin , author G. Pang , author Z. Zhang ,\ and\ author G. E. \ Karniadakis ,\ https://doi.org/10.1038/s42256-021-00302-5 journal journal Nat. Mach. Intell. \ volume 3 ,\ pages 218–229 ( year 2021 a ) NoStop

    work page doi:10.1038/s42256-021-00302-5 2021

  31. [31]

    Fourier Neural Operator for Parametric Partial Differential Equations

    author author Z. Li , author N. Kovachki , author K. Azizzadenesheli , author B. Liu , author K. Bhattacharya , author A. Stuart ,\ and\ author A. Anandkumar ,\ @noop \ note Eprint arXiv: 2010.08895 (2021) NoStop

    work page internal anchor Pith review Pith/arXiv arXiv 2010

  32. [32]

    Kovachki , author Z

    author author N. Kovachki , author Z. Li , author B. Liu , author K. Azizzadenesheli , author K. Bhattacharya , author A. Stuart ,\ and\ author A. Anandkumar ,\ https://doi.org/10.5555/3648699.3648788 journal journal J. Mach. Learn. Res. \ volume 24 ,\ pages 1--97 ( year 2023 ) NoStop

    work page doi:10.5555/3648699.3648788 2023

  33. [33]

    Guibas , author M

    author author J. Guibas , author M. Mardani , author Z. Li , author A. Tao , author A. Anandkumar ,\ and\ author B. Catanzaro ,\ @noop \ note Eprint arXiv: 2111.13587 (2021) NoStop

    work page arXiv 2021

  34. [34]

    Lu , author X

    author author L. Lu , author X. Meng , author Z. Mao ,\ and\ author G. E. \ Karniadakis ,\ https://doi.org/10.1137/19m1274067 journal journal SIAM Rev. \ volume 63 ,\ pages 208–228 ( year 2021 b ) NoStop

    work page doi:10.1137/19m1274067 2021

  35. [35]

    FourCastNet: A Global Data-driven High-resolution Weather Model using Adaptive Fourier Neural Operators

    author author J. Pathak , author S. Subramanian , author P. Harrington , author S. Raja , author A. Chattopadhyay , author M. Mardani , author T. Kurth , author D. Hall , author Z. Li , author K. Azizzadenesheli , author P. Hassanzadeh , author K. Kashinath ,\ and\ author A. Anandkumar ,\ @noop \ note Eprint arXiv: 2202.11214 (2022) NoStop

    work page internal anchor Pith review Pith/arXiv arXiv 2022

  36. [36]

    Jiang , author N

    author author P. Jiang , author N. Meinert , author H. Jordão , author C. Weisser , author S. Holgate , author A. Lavin , author B. L\" u tjens , author D. Newman , author H. Wainwright , author C. Walker ,\ and\ author P. Barnard ,\ @noop \ note Eprint arXiv: 2110.07100 (2021) NoStop

    work page arXiv 2021

  37. [37]

    Zhang , author C

    author author J. Zhang , author C. L. \ Benavides-Riveros ,\ and\ author L. Chen ,\ https://doi.org/10.1021/acs.jpclett.4c00598 journal journal J. Phys. Chem. Lett. \ volume 15 ,\ pages 3603--3610 ( year 2024 ) NoStop

    work page doi:10.1021/acs.jpclett.4c00598 2024

  38. [38]

    Ishizaki \ and\ author Y

    author author A. Ishizaki \ and\ author Y. Tanimura ,\ https://doi.org/10.1143/JPSJ.74.3131 journal journal J. Phys. Soc. Jpn. \ volume 74 ,\ pages 3131--3134 ( year 2005 ) NoStop

    work page doi:10.1143/jpsj.74.3131 2005

  39. [39]

    Hu , author R.-X

    author author J. Hu , author R.-X. \ Xu ,\ and\ author Y. Yan ,\ https://doi.org/10.1063/1.3484491 journal journal J. Chem. Phys. \ volume 133 ,\ pages 101106 ( year 2010 ) NoStop

    work page doi:10.1063/1.3484491 2010

  40. [40]

    Tanimura ,\ https://doi.org/10.1143/JPSJ.75.082001 journal journal J

    author author Y. Tanimura ,\ https://doi.org/10.1143/JPSJ.75.082001 journal journal J. Phys. Soc. Jpn. \ volume 75 ,\ pages 082001 ( year 2006 ) NoStop

    work page doi:10.1143/jpsj.75.082001 2006

  41. [41]

    Zhang \ and\ author Y

    author author J. Zhang \ and\ author Y. Tanimura ,\ https://doi.org/10.1063/5.0156264 journal journal J. Chem. Phys. \ volume 159 ,\ pages 014102 ( year 2023 ) NoStop

    work page doi:10.1063/5.0156264 2023

  42. [42]

    Kovachki , author S

    author author N. Kovachki , author S. Lanthaler ,\ and\ author S. Mishra ,\ https://doi.org/10.5555/3546258.3546548 journal journal J. Mach. Learn. Res. \ volume 22 ,\ pages 1--76 ( year 2021 ) NoStop

    work page doi:10.5555/3546258.3546548 2021

  43. [43]

    author author S. G. \ Rosofsky , author H. A. \ Majed ,\ and\ author E. A. \ Huerta ,\ https://doi.org/10.1088/2632-2153/acd168 journal journal Mach. learn.: Sci. Technol. \ volume 4 ,\ pages 025022 ( year 2023 ) NoStop

    work page doi:10.1088/2632-2153/acd168 2023

  44. [44]

    Adolphs \ and\ author T

    author author J. Adolphs \ and\ author T. Renger ,\ https://doi.org/10.1529/biophysj.105.079483 journal journal Biophys. J. \ volume 91 ,\ pages 2778--2797 ( year 2006 ) NoStop

    work page doi:10.1529/biophysj.105.079483 2006

  45. [45]

    Ishizaki \ and\ author G

    author author A. Ishizaki \ and\ author G. R. \ Fleming ,\ https://doi.org/10.1073/pnas.0908989106 journal journal Proc. Natl. Acad. Sci. \ volume 106 ,\ pages 17255--17260 ( year 2009 ) NoStop

    work page doi:10.1073/pnas.0908989106 2009

  46. [46]

    Shi , author L

    author author Q. Shi , author L. Chen , author G. Nan , author R.-X. \ Xu ,\ and\ author Y. Yan ,\ https://doi.org/10.1063/1.3077918 journal journal J. Chem. Phys. \ volume 130 ,\ pages 084105 ( year 2009 ) NoStop

    work page doi:10.1063/1.3077918 2009

  47. [47]

    \ Yeh \ and\ author S

    author author S.-H. \ Yeh \ and\ author S. Kais ,\ @noop journal journal J. Chem. Phys. \ volume 141 ( year 2014 ) NoStop

    work page 2014

  48. [48]

    Chen , author R

    author author L. Chen , author R. Zheng , author Y. Jing ,\ and\ author Q. Shi ,\ @noop journal journal J. Chem. Phys. \ volume 134 ( year 2011 ) NoStop

    work page 2011

  49. [49]

    Cho , author H

    author author M. Cho , author H. M. \ Vaswani , author T. Brixner , author J. Stenger ,\ and\ author G. R. \ Fleming ,\ https://doi.org/10.1021/jp050788d journal journal J. Phys. Chem. B \ volume 109 ,\ pages 10542–10556 ( year 2005 ) NoStop

    work page doi:10.1021/jp050788d 2005

  50. [50]

    Borrelli ,\ https://doi.org/10.1063/1.5099416 journal journal J

    author author R. Borrelli ,\ https://doi.org/10.1063/1.5099416 journal journal J. Chem. Phys. \ volume 150 ,\ pages 234102 ( year 2019 ) NoStop

    work page doi:10.1063/1.5099416 2019

  51. [51]

    Borrelli \ and\ author M

    author author R. Borrelli \ and\ author M. F. \ Gelin ,\ https://doi.org/10.1002/wcms.1539 journal journal WIREs Comput. Mol. Sci. \ volume 11 ,\ pages e1539 ( year 2021 ) NoStop

    work page doi:10.1002/wcms.1539 2021

  52. [52]

    Wang \ and\ author M

    author author H. Wang \ and\ author M. Thoss ,\ https://doi.org/10.1016/j.cplett.2004.03.052 journal journal Chem. Phys. Lett. \ volume 389 ,\ pages 43--50 ( year 2004 ) NoStop

    work page doi:10.1016/j.cplett.2004.03.052 2004

  53. [53]

    Wang \ and\ author M

    author author H. Wang \ and\ author M. Thoss ,\ https://doi.org/10.1016/j.chemphys.2007.12.004 journal journal Chem. Phys. \ volume 347 ,\ pages 139--151 ( year 2008 ) NoStop

    work page doi:10.1016/j.chemphys.2007.12.004 2007