pith. sign in

arxiv: 2511.03061 · v1 · submitted 2025-11-04 · ⚛️ physics.chem-ph · physics.comp-ph

Modal Backflow Neural Quantum States for Anharmonic Vibrational Calculations

Lexin Ding , Markus Reiher This is my paper

Pith reviewed 2026-05-18 00:51 UTC · model grok-4.3

classification ⚛️ physics.chem-ph physics.comp-ph
keywords neural quantum statesanharmonic vibrationsvibrational spectroscopymodal backflowquantum chemistryselected configurationbosonic ansatz
0
0 comments X

The pith

Modal backflow neural quantum states achieve spectroscopically accurate vibrational energies and transitions.

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

This paper develops a neural quantum state using modal backflow to solve the anharmonic vibrational Schrödinger equation for molecules. The design hardwires bosonic symmetry and particle conservation to make the network practical, sidestepping the expensive permanent calculations that plague standard backflow for identical bosons. It replaces Monte Carlo sampling with a selected-configuration scheme for precise observables and uses vibrational self-consistent field as pretraining to speed up and stabilize learning. When tested on artificial model Hamiltonians and real ab initio ones, the resulting wave functions yield zero-point energies and transition frequencies that match spectroscopic standards even in strongly anharmonic cases. A reader would care because vibrational spectroscopy relies on these quantities to interpret experimental spectra, and anharmonic effects are crucial for accurate predictions in molecular science.

Core claim

The modal backflow neural quantum state, equipped with a selected-configuration evaluation scheme and vibrational self-consistent field pretraining, is shown to produce zero-point energies and vibrational transitions of spectroscopic accuracy for anharmonic problems in both artificial and ab initio Hamiltonians.

What carries the argument

Modal backflow transformation, which applies a learned correction to the vibrational modes to capture anharmonic correlations while maintaining bosonic statistics.

If this is right

  • Accurate zero-point energies and transitions are obtained without stochastic sampling errors.
  • The method handles all degrees of anharmonicity, from weak to strong.
  • Pretraining stabilizes the variational optimization for high-accuracy targets.
  • It applies equally well to model potentials and those from electronic structure calculations.

Where Pith is reading between the lines

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

  • One could test the approach on larger polyatomic molecules to see if it scales better than traditional configuration interaction methods.
  • The selected-configuration idea might inspire similar deterministic schemes in electronic neural quantum states.
  • Extending this to time-dependent or finite-temperature vibrational problems could follow naturally from the stationary accuracy.

Load-bearing premise

The modal backflow transformation combined with the selected-configuration scheme can be optimized to deliver the claimed spectroscopic accuracy without hidden biases from the pretraining or configuration selection.

What would settle it

A direct comparison of the calculated vibrational transition frequencies for a well-studied anharmonic molecule such as H2O against high-level benchmark or experimental data; systematic deviations larger than a few wavenumbers would disprove the accuracy.

Figures

Figures reproduced from arXiv: 2511.03061 by Lexin Ding, Markus Reiher.

Figure 1
Figure 1. Figure 1: FIG. 1. (a) Architecture of the MBF network: ONVs of a set of bosonic modes are fed through [PITH_FULL_IMAGE:figures/full_fig_p009_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2. Distribution of the anharmonic correction of sampled anharmonic 4-mode Hamiltonians [PITH_FULL_IMAGE:figures/full_fig_p016_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: FIG. 3. (Top) Comparison of the optimization between FNN and MBF for targeting the ground [PITH_FULL_IMAGE:figures/full_fig_p017_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: FIG. 4. Error (cm [PITH_FULL_IMAGE:figures/full_fig_p018_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: FIG. 5. (Top) Optimization of the lowest eight vibrational levels (shifted by the ZPE) of ClO [PITH_FULL_IMAGE:figures/full_fig_p020_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: FIG. 6. Error (cm [PITH_FULL_IMAGE:figures/full_fig_p021_6.png] view at source ↗
read the original abstract

Neural quantum states (NQS) are a promising ansatz for solving many-body quantum problems due to their inherent expressiveness. Yet, this expressiveness can only be harnessed efficiently for treating identical particles if the suitable physical knowledge is hardwired into the neural network itself. For electronic structure, NQS based on backflow determinants has been shown to be a powerful ansatz for capturing strong correlation. By contrast, the analogue for bosons, backflow permanents, is unpractical due to the steep cost of computing the matrix permanent and due to the lack of particle conservation in common bosonic problems. To circumvent these obstacles, we introduce a modal backflow (MBF) NQS design and demonstrate its efficacy by solving the anharmonic vibrational problem. To accommodate the demand of high accuracy in spectroscopic calculations, we implement a selected-configuration scheme for evaluating physical observables and gradients, replacing the standard stochastic approach based on Monte Carlo sampling. A vibrational self-consistent field calculation is conveniently carried out within the MBF network, which serves as a pretraining step to accelerate and stabilize the optimization. In applications to both artificial and ab initio Hamiltonians, we find that the MBF network is capable of delivering spectroscopically accurate zero-point energies and vibrational transitions in all anharmonic regimes.

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

2 major / 2 minor

Summary. The manuscript introduces a modal backflow (MBF) neural quantum state ansatz for solving the anharmonic vibrational Schrödinger equation. The approach hardwires bosonic modal structure into the network to avoid the computational cost of permanents, replaces Monte Carlo sampling with a selected-configuration scheme for observables and gradients, and employs a VSCF calculation as pretraining. The central claim is that this MBF network yields spectroscopically accurate zero-point energies and vibrational transition frequencies (typically <1 cm⁻¹) for both artificial model Hamiltonians and ab initio potentials across all anharmonic regimes.

Significance. If the accuracy claims are substantiated, the work would represent a meaningful extension of neural quantum states to bosonic vibrational problems, providing a deterministic high-precision alternative to stochastic sampling methods. The combination of modal backflow with selected configurations and VSCF pretraining addresses practical barriers in applying NQS to spectroscopy and could scale to larger molecular systems where traditional variational or perturbative methods struggle with strong anharmonicity.

major comments (2)
  1. [Abstract / Selected-configuration scheme] Abstract and selected-configuration section: The claim of spectroscopic accuracy 'in all anharmonic regimes' for both artificial and ab initio Hamiltonians is load-bearing. The selected-configuration scheme (replacing Monte Carlo) must be shown to capture high-order mode couplings without systematic omission; otherwise the variational minimum and transition energies can be biased. The manuscript should include direct comparisons to exact diagonalization on small model Hamiltonians (e.g., 2–4 mode strongly anharmonic cases) with explicit error metrics and configuration-selection thresholds to demonstrate completeness.
  2. [Pretraining and optimization] Pretraining and optimization section: VSCF pretraining is used to accelerate and stabilize the MBF optimization. However, if the mean-field bias from VSCF is not fully escaped in strong anharmonicity, the final accuracy may be overstated. The paper should report convergence tests starting from different initializations or without pretraining, together with the magnitude of energy corrections achieved during the MBF optimization phase.
minor comments (2)
  1. [Abstract] The abstract states accuracy results but does not reference specific numerical tables, error bars, or baseline comparisons; these should be highlighted with section numbers in the abstract or introduction for immediate visibility.
  2. [Methods] Notation for the modal backflow transformation and the selected-configuration energy expression should be defined explicitly with equation numbers in the methods section to aid reproducibility.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their careful reading of the manuscript and for the constructive comments, which help strengthen the presentation of our results. We address each major comment below and have revised the manuscript accordingly.

read point-by-point responses

  1. Referee: [Abstract / Selected-configuration scheme] Abstract and selected-configuration section: The claim of spectroscopic accuracy 'in all anharmonic regimes' for both artificial and ab initio Hamiltonians is load-bearing. The selected-configuration scheme (replacing Monte Carlo) must be shown to capture high-order mode couplings without systematic omission; otherwise the variational minimum and transition energies can be biased. The manuscript should include direct comparisons to exact diagonalization on small model Hamiltonians (e.g., 2–4 mode strongly anharmonic cases) with explicit error metrics and configuration-selection thresholds to demonstrate completeness.

    Authors: We agree that explicit validation of the selected-configuration scheme against exact results is necessary to support the accuracy claims across anharmonic regimes. In the revised manuscript we have added direct comparisons to exact diagonalization for 2-, 3-, and 4-mode model Hamiltonians in strongly anharmonic regimes. These comparisons report absolute errors in zero-point energies and transition frequencies together with the configuration-selection thresholds employed. The results show that the selected configurations recover the dominant high-order couplings, with errors remaining below 1 cm⁻¹ and no evidence of systematic bias in the variational minimum. revision: yes

  2. Referee: [Pretraining and optimization] Pretraining and optimization section: VSCF pretraining is used to accelerate and stabilize the MBF optimization. However, if the mean-field bias from VSCF is not fully escaped in strong anharmonicity, the final accuracy may be overstated. The paper should report convergence tests starting from different initializations or without pretraining, together with the magnitude of energy corrections achieved during the MBF optimization phase.

    Authors: We acknowledge the importance of demonstrating that VSCF pretraining does not leave a residual mean-field bias. In the revised manuscript we have included additional optimization trajectories started from random initializations without pretraining. These tests confirm that the MBF network escapes the mean-field solution, producing energy corrections of several hundred cm⁻¹ in the strongest anharmonicity cases. We report the magnitude of these corrections and the associated convergence behavior to show that the final accuracies are not overstated. revision: yes

Circularity Check

0 steps flagged

No significant circularity: MBF ansatz and selected-configuration evaluation are independent of target observables

full rationale

The paper introduces a modal backflow transformation for bosonic vibrational NQS, replacing permanents with a modal representation and adopting a deterministic selected-configuration scheme plus VSCF pretraining. These are presented as architectural and algorithmic choices that enable optimization, not as redefinitions of the zero-point energy or transition frequencies. No equation reduces the claimed spectroscopic accuracy to a fitted parameter defined by the same network; results are obtained by variational minimization on external Hamiltonians. Self-citations to prior NQS work are non-load-bearing and do not substitute for the new bosonic construction or the numerical benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Review based on abstract only; no explicit free parameters, axioms, or invented entities are stated. The neural network architecture and modal backflow transformation are implicit but not quantified.

pith-pipeline@v0.9.0 · 5753 in / 1099 out tokens · 56069 ms · 2026-05-18T00:51:58.203371+00:00 · methodology

discussion (0)

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

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

97 extracted references · 97 canonical work pages · 1 internal anchor

  1. [1]

    ,0), or another seed state of choice (depending on the level of excitation)

    Initialize Ψ θ(0) and V (0) as {n0} where n0 = (0 , 0, . . . ,0), or another seed state of choice (depending on the level of excitation)

    work page

  2. [2]

    In step t, compute the set V ′ = ˆHV (t) consisting of configurations connected by ˆH to V (t)

    work page

  3. [3]

    Select Ns configurations from V (t) ∪ V ′ with the highest amplitudes |Ψθ(n)| to form the new set of selected states V (t+1)

    work page

  4. [4]

    Evaluate the energy gradients ∂E/∂θ m and update the network parameters θ(t) ← θ(t+1)

    work page

  5. [5]

    Repeat from step 2 until the maximum number of iterations is reached. The above scheme was first implemented to solve the electronic Hamiltonian.[77] In our case, the vibrational Hamiltonian contains long strings of creation and annihilation operators that go beyond the two-body terms. The implication is that the connected subspace V ′ in the vibrational ...

    work page

  6. [6]

    (29) The default hyperparameters used in this work are η = 0.05, βa 1 = 0.9, βb 1 = 0.5, βc 1 = 100, β2 = 0.99

    exp " − t − 1 βc 1 2# . (29) The default hyperparameters used in this work are η = 0.05, βa 1 = 0.9, βb 1 = 0.5, βc 1 = 100, β2 = 0.99. VSCF pretraining. Pretraining the network parameters is a proven technique for accelerating and stabilizing the optimization of the network.[32, 62] Here, we can pretrain the backflow-free part ϕ(0) i of the ONV-dependent...

    work page

  7. [7]

    x77p37232o2Tu1mrTwrsB7rg170=

    and that of ˆHvib. In Figure 2 we plotted the distribution anharmonic correction of 103 sampled Hamiltonian, with κ(3) ijk and κ(4) ijkl both set to represent weak, mod- erate, and strong anharmonicity. For each Hamiltonian, the ground state was calculated with exact diagonalization and Nmax was set to 9. We found that the three settings indeed 15 FIG. 2....

    work page 2000

  8. [8]

    K. D. Vogiatzis, D. Ma, J. Olsen, L. Gagliardi, and W. A. De Jong, Pushing configuration- interaction to the limit: Towards massively parallel MCSCF calculations, J. Chem. Phys.147, 184111 (2017)

    work page 2017

  9. [9]

    H. Gao, S. Imamura, A. Kasagi, and E. Yoshida, Distributed implementation of full configu- ration interaction for one trillion determinants, J. Chem. Theory Comput. 20, 1185 (2024)

    work page 2024

  10. [10]

    C. F. Bender and E. R. Davidson, Studies in configuration interaction: The first-row diatomic hydrides, Phys. Rev. 183, 23 (1969). 25

    work page 1969

  11. [11]

    J. B. Schriber and F. A. Evangelista, Communication: An adaptive configuration interaction approach for strongly correlated electrons with tunable accuracy, J. Chem. Phys. 144, 161106 (2016)

    work page 2016

  12. [12]

    A. A. Holmes, N. M. Tubman, and C. Umrigar, Heat-bath configuration interaction: An efficient selected configuration interaction algorithm inspired by heat-bath sampling, J. Chem. Theory Comput. 12, 3674 (2016)

    work page 2016

  13. [13]

    Liu and M

    W. Liu and M. R. Hoffmann, iCI: Iterative CI toward full CI, J. Chem. Theory Comput. 12, 1169 (2016)

    work page 2016

  14. [14]

    P. M. Zimmerman, Incremental full configuration interaction, J. Chem. Phys. 146, 224104 (2017)

    work page 2017

  15. [15]

    V. G. Chilkuri and F. Neese, Comparison of many-particle representations for selected con- figuration interaction: II. Numerical benchmark calculations, J. Chem. Theory Comput. 17, 2868 (2021)

    work page 2021

  16. [16]

    S. R. White, Density matrix formulation for quantum renormalization groups, Phys. Rev. Lett. 69, 2863 (1992)

    work page 1992

  17. [17]

    Baiardi and M

    A. Baiardi and M. Reiher, The density matrix renormalization group in chemistry and molec- ular physics: Recent developments and new challenges, J. Chem. Phys. 152, 040903 (2020)

    work page 2020

  18. [18]

    H. Ma, U. Schollw¨ ock, and Z. Shuai,Density Matrix Renormalization Group (DMRG)-Based Approaches in Computational Chemistry (Elsevier, 2022)

    work page 2022

  19. [19]

    G. H. Booth, A. J. Thom, and A. Alavi, Fermion Monte Carlo without fixed nodes: A game of life, death, and annihilation in Slater determinant space, J. Chem. Phys. 131, 054106 (2009)

    work page 2009

  20. [20]

    J. J. Eriksen, The shape of full configuration interaction to come, J. Phys. Chem. Lett. 12, 418 (2020)

    work page 2020

  21. [21]

    Zhang and H

    S. Zhang and H. Krakauer, Quantum Monte Carlo Method using Phase-Free Random Walks with Slater Determinants, Phys. Rev. Lett. 90, 136401 (2003)

    work page 2003

  22. [22]

    J. Lee, H. Q. Pham, and D. R. Reichman, Twenty years of auxiliary-field quantum Monte Carlo in quantum chemistry: An overview and assessment on main group chemistry and bond-breaking, J. Chem. Theory Comput. 18, 7024 (2022)

    work page 2022

  23. [23]

    Carleo and M

    G. Carleo and M. Troyer, Solving the quantum many-body problem with artificial neural networks, Science 355, 602 (2017). 26

    work page 2017

  24. [24]

    D. E. Rumelhart, G. E. Hinton, and R. J. Williams, Learning representations by back- propagating errors, Nature 323, 533 (1986)

    work page 1986

  25. [25]

    LeCun, B

    Y. LeCun, B. Boser, J. S. Denker, D. Henderson, R. E. Howard, W. Hubbard, and L. D. Jackel, Backpropagation applied to handwritten zip code recognition, Neural Comput. 1, 541 (1989)

    work page 1989

  26. [26]

    D. F. Specht, A general regression neural network, IEEE Trans. Neural Netw. Learn. Syst. 2, 568 (1991)

    work page 1991

  27. [27]

    Robbins and S

    H. Robbins and S. Monro, A stochastic approximation method, Ann. Math. Stat 22, 400 (1951)

    work page 1951

  28. [28]

    B. T. Polyak, Some methods of speeding up the convergence of iteration methods, Zh. Vych. Mat. 4, 1 (1964)

    work page 1964

  29. [29]

    Duchi, E

    J. Duchi, E. Hazan, and Y. Singer, Adaptive subgradient methods for online learning and stochastic optimization, J. Mach. Learn. Res. 12, 2121 (2011)

    work page 2011

  30. [30]

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

    work page internal anchor Pith review Pith/arXiv arXiv 2014

  31. [31]

    Cybenko, Approximation by superpositions of a sigmoidal function, Math

    G. Cybenko, Approximation by superpositions of a sigmoidal function, Math. Control Signals Syst. 2, 303 (1989)

    work page 1989

  32. [32]

    Hornik, M

    K. Hornik, M. Stinchcombe, and H. White, Multilayer feedforward networks are universal approximators, Neural Netw. 2, 359 (1989)

    work page 1989

  33. [33]

    Sharir, A

    O. Sharir, A. Shashua, and G. Carleo, Neural tensor contractions and the expressive power of deep neural quantum states, Phys. Rev. B 106, 205136 (2022)

    work page 2022

  34. [34]

    D.-L. Deng, X. Li, and S. Das Sarma, Quantum entanglement in neural network states, Phys. Rev. X 7, 021021 (2017)

    work page 2017

  35. [35]

    Can Neural Quantum States Learn Volume-Law Ground States?

    Z. Denis, A. Sinibaldi, and G. Carleo, Comment on “Can Neural Quantum States Learn Volume-Law Ground States?”, Phys. Rev. Lett. 134, 079701 (2025)

    work page 2025

  36. [36]

    Luo and B

    D. Luo and B. K. Clark, Backflow transformations via neural networks for quantum many- body wave functions, Phys. Rev. Lett. 122, 226401 (2019)

    work page 2019

  37. [37]

    Manzhos, Machine learning for the solution of the Schr¨ odinger equation, Mach

    S. Manzhos, Machine learning for the solution of the Schr¨ odinger equation, Mach. Learn.: Sci. Technol. 1, 013002 (2020)

    work page 2020

  38. [38]

    Hermann, Z

    J. Hermann, Z. Sch¨ atzle, and F. No´ e, Deep-neural-network solution of the electronic schr¨ odinger equation, Nat. Chem.12, 891 (2020). 27

    work page 2020

  39. [39]

    D. Pfau, J. S. Spencer, A. G. D. G. Matthews, and W. M. C. Foulkes, Ab initio solution of the many-electron Schr¨ odinger equation with deep neural networks, Phys. Rev. Res.2, 033429 (2020)

    work page 2020

  40. [40]

    Robledo Moreno, G

    J. Robledo Moreno, G. Carleo, A. Georges, and J. Stokes, Fermionic wave functions from neural-network constrained hidden states, Proc. Natl. Acad. Sci. 119, e2122059119 (2022)

    work page 2022

  41. [41]

    Scherbela, R

    M. Scherbela, R. Reisenhofer, L. Gerard, P. Marquetand, and P. Grohs, Solving the elec- tronic Schr¨ odinger equation for multiple nuclear geometries with weight-sharing deep neural networks, Nat. Comput. Sci. 2, 331 (2022)

    work page 2022

  42. [42]

    Hermann, J

    J. Hermann, J. Spencer, K. Choo, A. Mezzacapo, W. M. C. Foulkes, D. Pfau, G. Carleo, and F. No´ e, Ab initio quantum chemistry with neural-network wavefunctions, Nat. Rev. Chem.7, 692 (2023)

    work page 2023

  43. [43]

    T. Zhao, J. Stokes, and S. Veerapaneni, Scalable neural quantum states architecture for quan- tum chemistry, Mach. Learn.: Sci. Technol. 4, 025034 (2023)

    work page 2023

  44. [44]

    K. Choo, A. Mezzacapo, and G. Carleo, Fermionic neural-network states for ab-initio electronic structure, Nat. Comm. 11, 2368 (2020)

    work page 2020

  45. [45]

    Yoshioka, W

    N. Yoshioka, W. Mizukami, and F. Nori, Solving quasiparticle band spectra of real solids using neural-network quantum states, Comm. Phys. 4, 106 (2021)

    work page 2021

  46. [46]

    Zhang, R.-S

    Q. Zhang, R.-S. Wang, and L. Wang, Neural canonical transformations for vibrational spectra of molecules, J. Chem. Phys. 161, 024103 (2024)

    work page 2024

  47. [47]

    J. M. Bowman, K. Christoffel, and F. Tobin, Application of SCF-SI theory to vibrational motion in polyatomic molecules, J. Phys. Chem. 83, 905 (1979)

    work page 1979

  48. [48]

    Garnier, M

    R. Garnier, M. Odunlami, V. Le Bris, D. B´ egu´ e, I. Baraille, and O. Coulaud, Adaptive vibra- tional configuration interaction (A-VCI): A posteriori error estimation to efficiently compute anharmonic IR spectra, J. Chem. Phys. 144, 204123 (2016)

    work page 2016

  49. [49]

    Schr¨ oder and G

    B. Schr¨ oder and G. Rauhut, Incremental vibrational configuration interaction theory, iVCI: Implementation and benchmark calculations, J. Chem. Phys. 154, 124114 (2021)

    work page 2021

  50. [50]

    Schr¨ oder and G

    B. Schr¨ oder and G. Rauhut, Vibrational configuration interaction theory, in Vibrational Dy- namics of Molecules (World Scientific Publishing Company, 2022) Chap. 1, pp. 1–40

    work page 2022

  51. [51]

    Christiansen, Vibrational coupled cluster theory, J

    O. Christiansen, Vibrational coupled cluster theory, J. Chem. Phys. 120, 2149 (2004)

    work page 2004

  52. [52]

    J. A. Faucheaux and S. Hirata, Higher-order diagrammatic vibrational coupled-cluster theory, J. Chem. Phys. 143, 134105 (2015). 28

    work page 2015

  53. [53]

    N. K. Madsen, I. H. Godtliebsen, S. A. Losilla, and O. Christiansen, Tensor-decomposed vibrational coupled-cluster theory: Enabling large-scale, highly accurate vibrational-structure calculations, The Journal of Chemical Physics 148, 024103 (2018)

    work page 2018

  54. [54]

    Christiansen, Vibrational coupled cluster theory, in Vibrational Dynamics of Molecules (World Scientific Publishing Company, 2022) Chap

    O. Christiansen, Vibrational coupled cluster theory, in Vibrational Dynamics of Molecules (World Scientific Publishing Company, 2022) Chap. 2, pp. 41–79

    work page 2022

  55. [55]

    Baiardi, C

    A. Baiardi, C. J. Stein, V. Barone, and M. Reiher, Vibrational density matrix renormalization group, J. Chem. Theory Comput. 13, 3764 (2017)

    work page 2017

  56. [56]

    Baiardi, C

    A. Baiardi, C. J. Stein, V. Barone, and M. Reiher, Optimization of highly excited matrix product states with an application to vibrational spectroscopy, J. Chem. Phys. 150, 094113 (2019)

    work page 2019

  57. [57]

    H. R. Larsson, Computing vibrational eigenstates with tree tensor network states (TTNS), J. Chem, Phys. 151, 204102 (2019)

    work page 2019

  58. [58]

    Glaser, A

    N. Glaser, A. Baiardi, and M. Reiher, Tensor network states for vibrational spectroscopy, in Vibrational Dynamics of Molecules (World Scientific Publishing Company, 2022) Chap. 3, pp. 80–144

    work page 2022

  59. [59]

    H. R. Larsson, Benchmarking vibrational spectra: 5000 accurate eigenstates of acetonitrile using tree tensor network states, J. Phys. Chem. Lett. 16, 3991 (2025)

    work page 2025

  60. [60]

    Puzzarini, J

    C. Puzzarini, J. Bloino, N. Tasinato, and V. Barone, Accuracy and interpretability: The devil and the holy grail. New routes across old boundaries in computational spectroscopy, Chem. Rev. 119, 8131 (2019)

    work page 2019

  61. [61]

    Wang and M

    H. Wang and M. Thoss, Multilayer formulation of the multiconfiguration time-dependent Hartree theory, J. Chem. Phys. 119, 1289 (2003)

    work page 2003

  62. [62]

    Manthe, A multilayer multiconfigurational time-dependent Hartree approach for quantum dynamics on general potential energy surfaces, J

    U. Manthe, A multilayer multiconfigurational time-dependent Hartree approach for quantum dynamics on general potential energy surfaces, J. Chem. Phys. 128, 164116 (2008)

    work page 2008

  63. [63]

    Vendrell and H.-D

    O. Vendrell and H.-D. Meyer, Multilayer multiconfiguration time-dependent Hartree method: Implementation and applications to a Henon–Heiles Hamiltonian and to pyrazine, J. Chem. Phys. 134, 044135 (2011)

    work page 2011

  64. [64]

    Glaser, A

    N. Glaser, A. Baiardi, and M. Reiher, Flexible DMRG-based framework for anharmonic vi- brational calculations, J. Chem. Theory Comput. 19, 9329 (2023)

    work page 2023

  65. [65]

    R. L. Fulton and M. Gouterman, Vibronic coupling. I. Mathematical treatment for two elec- tronic states, J. Chem. Phys. 35, 1059 (1961). 29

    work page 1961

  66. [66]

    Meng and H.-D

    Q. Meng and H.-D. Meyer, A multilayer MCTDH study on the full dimensional vibronic dynamics of naphthalene and anthracene cations, J. Chem. Phys. 138, 014313 (2013)

    work page 2013

  67. [67]

    D. S. Wang and S. F. Yelin, A roadmap toward the theory of vibrational polariton chemistry, ACS Photonics 8, 2818 (2021)

    work page 2021

  68. [68]

    A. D. Dunkelberger, B. S. Simpkins, I. Vurgaftman, and J. C. Owrutsky, Vibration-cavity polariton chemistry and dynamics, Annu. Rev. Phys. Chem. 73, 429 (2022)

    work page 2022

  69. [69]

    Liu and B

    A.-J. Liu and B. K. Clark, Neural network backflow for ab initio quantum chemistry, Phys. Rev. B 110, 115137 (2024)

    work page 2024

  70. [70]

    Passetti, D

    G. Passetti, D. Hofmann, P. Neitemeier, L. Grunwald, M. A. Sentef, and D. M. Kennes, Can neural quantum states learn volume-law ground states?, Phys. Rev. Lett. 131, 036502 (2023)

    work page 2023

  71. [71]

    J. K. Watson, Simplification of the molecular vibration-rotation hamiltonian, Mol. Phys. 15, 479 (1968)

    work page 1968

  72. [72]

    Carbonniere and V

    P. Carbonniere and V. Barone, Coriolis couplings in variational computations of vibrational spectra beyond the harmonic approximation: implementation and validation, Chem. Phys. Lett. 392, 365 (2004)

    work page 2004

  73. [73]

    R. P. Feynman and M. Cohen, Energy spectrum of the excitations in liquid helium, Phys. Rev. 102, 1189 (1956)

    work page 1956

  74. [74]

    Y. Kwon, D. M. Ceperley, and R. M. Martin, Effects of three-body and backflow correlations in the two-dimensional electron gas, Phys. Rev. B 48, 12037 (1993)

    work page 1993

  75. [75]

    L. F. Tocchio, F. Becca, A. Parola, and S. Sorella, Role of backflow correlations for the nonmagnetic phase of the t–t ′ hubbard model, Phys. Rev. B 78, 041101 (2008)

    work page 2008

  76. [76]

    L. G. Valiant, The complexity of computing the permanent, Theor. Comput. Sci.8, 189 (1979)

    work page 1979

  77. [77]

    Christiansen, A second quantization formulation of multimode dynamics, J

    O. Christiansen, A second quantization formulation of multimode dynamics, J. Chem. Phys. 120, 2140 (2004)

    work page 2004

  78. [78]

    J. M. Bowman, The self-consistent-field approach to polyatomic vibrations, Acc. Chem. Res. 19, 202 (1986)

    work page 1986

  79. [79]

    Gerber and M

    R. Gerber and M. A. Ratner, Self-consistent-field methods for vibrational excitations in poly- atomic systems, Adv. Chem. Phys. 70, 97 (1988)

    work page 1988

  80. [80]

    Carleo, K

    G. Carleo, K. Choo, D. Hofmann, J. E. T. Smith, T. Westerhout, F. Alet, E. J. Davis, S. Efthymiou, I. Glasser, S.-H. Lin, M. Mauri, G. Mazzola, C. B. Mendl, E. van Nieuwenburg, O. O’Reilly, H. Th´ eveniaut, G. Torlai, F. Vicentini, and A. Wietek, Netket: A machine learning 30 toolkit for many-body quantum systems, SoftwareX , 100311 (2019)

    work page 2019

Showing first 80 references.