pith. sign in

arxiv: 2512.21772 · v2 · submitted 2025-12-25 · ❄️ cond-mat.str-el · cond-mat.stat-mech

Time-dependent fluctuating local field approach for description of the correlated fermions dynamics

L.D. Silakov , Ya.S. Lyakhova , A.N. Rubtsov This is my paper

Pith reviewed 2026-05-16 19:16 UTC · model grok-4.3

classification ❄️ cond-mat.str-el cond-mat.stat-mech
keywords time-dependent fluctuating local fieldHubbard modelcorrelated fermionsnon-equilibrium dynamicsmean-field theoryexact diagonalizationreduced basis
0
0 comments X

The pith

Time-dependent fluctuating local fields reduce the Schrödinger equation for correlated fermions to a small generalized eigenvalue problem.

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

The paper develops a time-dependent version of the fluctuating local field method to model the dynamics of interacting fermions. It approximates the full many-body wavefunction as an ensemble of simpler non-interacting states, each feeling a classical field whose time-dependent probability distribution carries the effects of correlations. This turns the time-dependent Schrödinger equation into a generalized eigenvalue problem in a much smaller basis. Tests on half-filled two-dimensional Hubbard lattices show results that outperform mean-field theory and match exact diagonalization for the frequencies and amplitudes of oscillations. The low computational cost makes the approach practical for driven systems that are otherwise hard to simulate.

Core claim

The TD-FLF method approximates the wavefunction as an ensemble of non-interacting states subject to a classical fluctuating field, with dynamics encoded in the field's time-dependent distribution. This reduces the time-dependent Schrödinger equation to a generalized eigenvalue problem in a significantly reduced basis. Applied to half-filled 2D Hubbard lattices, TD-FLF yields highly accurate results, outperforming mean-field theory and capturing oscillation frequencies and amplitudes in good agreement with exact diagonalization.

What carries the argument

The time-dependent distribution of a classical fluctuating local field that encodes correlations and dynamics within an ensemble of non-interacting fermion states.

If this is right

  • TD-FLF outperforms mean-field theory for time-dependent observables in the Hubbard model.
  • It reproduces oscillation frequencies and amplitudes in agreement with exact diagonalization.
  • The method has low computational cost and can handle driven correlated systems.
  • It extends the stationary fluctuating local field approach to time-dependent problems.

Where Pith is reading between the lines

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

  • The method could be tested on other lattice geometries or interaction strengths where exact diagonalization becomes intractable.
  • Efficient ways to represent or sample the field distribution might allow application to three-dimensional or larger systems.
  • The reduced-basis eigenvalue structure suggests possible hybrids with tensor-network or stochastic techniques for non-equilibrium dynamics.

Load-bearing premise

The many-body wavefunction can be accurately approximated as an ensemble of non-interacting states subject to a classical fluctuating field whose time-dependent distribution encodes the dynamics.

What would settle it

Exact diagonalization on a small half-filled 2D Hubbard lattice showing large mismatches in time-dependent oscillation frequencies or amplitudes compared to TD-FLF predictions.

Figures

Figures reproduced from arXiv: 2512.21772 by A.N. Rubtsov, L.D. Silakov, Ya.S. Lyakhova.

Figure 1
Figure 1. Figure 1: Spectrum of matrix 1 eigenvalues λ divided by the maximum eigenvalue λmax and the size n0 of effective basis for 2 × 2 and 2 × 4 half-filled Hubbard lattices. such approximation we calculate the ground state energy of two-dimensional Hubbard lattice of different sizes and at different values of U. The results, obtained within FLF and its parental MF approximation, were compared to ED data, and are listed i… view at source ↗
Figure 3
Figure 3. Figure 3: Evolution of magnetization of 2 × 4 half-filled Hub￾bard lattice embedded in h = 0.5 magnetic field. MF approx￾imation and TD-FLF results are compared with the numeri￾cally exact (Exact) reference data. First, we see that the larger U is, the worse results gives MF approximation. TD-FLF in contrast catches changes is the temporal fluctuation pattern of the magnetization [PITH_FULL_IMAGE:figures/full_fig_p… view at source ↗
Figure 4
Figure 4. Figure 4: Fourier spectrum of temporal magnetization oscil [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗
read the original abstract

We formulate a time-dependent Fluctuating Local Field (TD-FLF) method for correlated fermion dynamics, extending the stationary FLF approach. The wavefunction is approximated as an ensemble of non-interacting states subject to a classical fluctuating field, with dynamics encoded in the field's time-dependent distribution. This reduces the time-dependent Schr\"odinger equation to a generalized eigenvalue problem in a significantly reduced basis. Applied to half-filled 2D Hubbard lattices, TD-FLF yields highly accurate results, outperforming mean-field theory and capturing oscillation frequencies and amplitudes in good agreement with exact diagonalization. Its low computational cost and flexibility make TD-FLF a promising tool for simulating driven correlated systems.

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 the time-dependent Fluctuating Local Field (TD-FLF) method as an extension of the stationary FLF approach for simulating dynamics of correlated fermions. The many-body wavefunction is approximated as an ensemble of non-interacting Slater determinants subject to a classical fluctuating local field whose time-dependent distribution encodes the dynamics. This reduces the time-dependent Schrödinger equation to a generalized eigenvalue problem in a reduced basis. When applied to half-filled 2D Hubbard lattices, the method is reported to produce results in good agreement with exact diagonalization, outperforming mean-field theory in reproducing oscillation frequencies and amplitudes, while offering low computational cost for driven systems.

Significance. If the accuracy claims hold with quantitative support, TD-FLF would provide a practical, scalable tool for nonequilibrium simulations of strongly correlated electrons on lattices larger than those accessible to exact diagonalization. It could bridge mean-field and exact methods for driven Hubbard-like models, with potential applications in ultrafast spectroscopy and quantum materials dynamics.

major comments (2)
  1. [Abstract and Results section] The central claim of high accuracy and agreement with exact diagonalization on half-filled 2D Hubbard models (abstract) lacks any quantitative error metrics, system sizes, interaction strengths (e.g., U/t values), or explicit comparisons in the provided description. This makes it impossible to evaluate whether the method truly outperforms mean-field while matching ED frequencies and amplitudes.
  2. [Method derivation (TDSE reduction)] The reduction of the TDSE to a generalized eigenvalue problem in the non-interacting basis (described in the method) assumes that the time-dependent classical field distribution fully encodes all correlation effects and entanglement without truncation or bias. In the strongly correlated regime (U/t ≳ 4), where Mott physics and antiferromagnetic correlations appear, this classical-field ensemble may miss higher-order cumulants, undermining the claim of capturing dynamics beyond mean-field.
minor comments (2)
  1. [Abstract] The abstract would be strengthened by including at least one concrete benchmark (e.g., lattice size, U/t, and a numerical error measure) to support the accuracy statements.
  2. [Method section] Notation for the fluctuating field distribution and its evolution rule should be defined more explicitly to clarify how the ensemble is generated at each time step.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive report and detailed comments. We address each major point below. Where the comments identify missing quantitative support or insufficient discussion of limitations, we have revised the manuscript accordingly.

read point-by-point responses

  1. Referee: [Abstract and Results section] The central claim of high accuracy and agreement with exact diagonalization on half-filled 2D Hubbard models (abstract) lacks any quantitative error metrics, system sizes, interaction strengths (e.g., U/t values), or explicit comparisons in the provided description. This makes it impossible to evaluate whether the method truly outperforms mean-field while matching ED frequencies and amplitudes.

    Authors: We agree that quantitative metrics are essential for evaluating the claims. In the revised manuscript we have added explicit error metrics (RMS deviations in oscillation frequencies and amplitudes relative to ED), specified the lattice sizes used (primarily 4x4 with selected 6x6 checks), interaction strengths (U/t = 4, 6, 8), and included direct side-by-side comparisons with both ED and mean-field results in new tables and figures. These additions make the performance claims verifiable. revision: yes

  2. Referee: [Method derivation (TDSE reduction)] The reduction of the TDSE to a generalized eigenvalue problem in the non-interacting basis (described in the method) assumes that the time-dependent classical field distribution fully encodes all correlation effects and entanglement without truncation or bias. In the strongly correlated regime (U/t ≳ 4), where Mott physics and antiferromagnetic correlations appear, this classical-field ensemble may miss higher-order cumulants, undermining the claim of capturing dynamics beyond mean-field.

    Authors: The TD-FLF ansatz is exact within the chosen ensemble of Slater determinants whose time-dependent field distribution is determined self-consistently from the equations of motion for the one-body density matrix. This construction goes beyond static mean-field by allowing the field distribution to fluctuate and thereby generate dynamic correlations. We acknowledge that the classical-field representation cannot capture all higher-order cumulants or full entanglement structure in the deep Mott regime. We have therefore added a dedicated paragraph in the revised manuscript discussing the range of validity, the moments reproduced by the distribution, and direct comparisons with ED that quantify the residual deviations for U/t ≳ 6. revision: partial

Circularity Check

0 steps flagged

TD-FLF derivation is self-contained; no load-bearing reductions to inputs or self-citations

full rationale

The paper frames TD-FLF as an independent approximation: the many-body wavefunction is approximated as an ensemble of non-interacting Slater determinants driven by a classical fluctuating field whose time-dependent distribution is evolved from the TDSE. This directly yields a generalized eigenvalue problem in the reduced basis without any fitted parameters being relabeled as predictions, without self-citations invoked as uniqueness theorems, and without ansatzes smuggled from prior author work. The reported agreement with exact diagonalization on half-filled 2D Hubbard lattices is presented as a numerical validation of the approximation, not a tautological consequence of the method's construction. No steps in the provided abstract or description exhibit self-definitional loops or renaming of known results.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Based solely on the abstract, the central claim rests on the validity of the ensemble approximation with a classical fluctuating field; no explicit free parameters, additional axioms, or invented entities are named.

pith-pipeline@v0.9.0 · 5419 in / 1060 out tokens · 21952 ms · 2026-05-16T19:16:32.695337+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

26 extracted references · 26 canonical work pages

  1. [1]

    M. Ma, X. Zhang, X. Chen, H. Xiong, L. Xu, T. Cheng, J. Yuan, and B. Shen, Nat Commun14, 7142 (2023)

    work page 2023

  2. [2]

    X. Wang, C. Lu, and M. Yang, Sci Rep10, 8539 (2020)

    work page 2020

  3. [3]

    M. Köhl, H. Moritz, T. Stöferle, K. Günter, and T. Esslinger, Phys. Rev. Lett.94, 080403 (2005), URL https://link.aps.org/doi/10.1103/PhysRevLett.94. 080403

    work page doi:10.1103/physrevlett.94 2005

  4. [4]

    Sponselee, L

    K. Sponselee, L. Freystatzky, B. Abeln, M. Diem, B. Hundt, A. Kochanke, T. Ponath, B. Santra, L. Mathey, K. Sengstock, et al., Quantum Science and Technology 4, 014002 (2018), URLhttps://dx.doi.org/10.1088/ 2058-9565/aadccd

    work page 2018

  5. [5]

    P. J. Ollitrault, A. Miessen, and I. Tavernelli, Accounts of Chemical Research54, 4229 (2021), pMID: 34787398, https://doi.org/10.1021/acs.accounts.1c00514, URL https://doi.org/10.1021/acs.accounts.1c00514

    work page doi:10.1021/acs.accounts.1c00514 2021

  6. [6]

    Jafarizadeh, F

    A. Jafarizadeh, F. Pollmann, and A. Gammon-Smith, Phys. Rev. Res.7, 043018 (2025), URLhttps://link. aps.org/doi/10.1103/PhysRevResearch.7.043018

    work page doi:10.1103/physrevresearch.7.043018 2025

  7. [7]

    I. Vona, M. Mestyán, and B. Pozsgay, Phys. Rev. B 111, 144306 (2025), URLhttps://link.aps.org/doi/ 10.1103/PhysRevB.111.144306

    work page doi:10.1103/physrevb.111.144306 2025

  8. [8]

    Stroeks, D

    M. Stroeks, D. Lenterman, B. M. Terhal, and Y. Herasy- menko, Phys. Rev. Res.7, 043176 (2025), URLhttps: //link.aps.org/doi/10.1103/zmwm-gdmw

    work page doi:10.1103/zmwm-gdmw 2025

  9. [9]

    Navickas, R

    T. Navickas, R. J. MacDonell, C. H. Valahu, V. C. Olaya- Agudelo, F. Scuccimarra, M. J. Millican, V. G. Matsos, H. L. Nourse, A. D. Rao, M. J. Biercuk, et al., Journal of the American Chemical Society147, 23566 (2025)

    work page 2025

  10. [10]

    F. Alam, J. L. Bosse, I. Čepait˙ e, A. Chapman, L. Clinton, M. Crichigno, E. Crosson, T. Cubitt, C. Derby, O. Dow- inton, et al.,Fermionic dynamics on a trapped-ion quan- tum computer beyond exact classical simulation(2025), 2510.26300, URLhttps://arxiv.org/abs/2510.26300

    work page arXiv 2025

  11. [11]

    Montorsi and V

    A. Montorsi and V. Penna, Phys. Rev. B55, 8226 (1997), URLhttps://link.aps.org/doi/10.1103/ PhysRevB.55.8226

    work page 1997

  12. [12]

    Schiró and M

    M. Schiró and M. Fabrizio, Phys. Rev. B83, 165105 (2011), URLhttps://link.aps.org/doi/10. 1103/PhysRevB.83.165105

    work page 2011

  13. [13]

    N. T. Maitra, Annual Review of Physical Chem- istry73, 117 (2022), ISSN 1545-1593, URL https://www.annualreviews.org/content/journals/ 10.1146/annurev-physchem-082720-124933

    work page doi:10.1146/annurev-physchem-082720-124933 2022

  14. [14]

    Karlsson, A

    D. Karlsson, A. Privitera, and C. Verdozzi, Phys. Rev. Lett.106, 116401 (2011), URLhttps://link.aps.org/ doi/10.1103/PhysRevLett.106.116401

    work page doi:10.1103/physrevlett.106.116401 2011

  15. [15]

    H. T. M. Nghiem and T. A. Costi, Phys. Rev. B 90, 035129(2014), URLhttps://link.aps.org/doi/10. 1103/PhysRevB.90.035129

    work page 2014

  16. [16]

    M. S. Church and B. M. Rubenstein, J. Chem. Phys.154, 184103 (2021)

    work page 2021

  17. [17]

    Caillat, J

    J. Caillat, J. Zanghellini, M. Kitzler, O. Koch, W. Kreuzer, and A. Scrinzi, Phys. Rev. A71, 012712 (2005), URLhttps://link.aps.org/doi/10. 1103/PhysRevA.71.012712

    work page 2005

  18. [18]

    D. J. Haxton, K. V. Lawler, and C. W. McCurdy, Phys. Rev. A83, 063416 (2011), URLhttps://link.aps.org/ doi/10.1103/PhysRevA.83.063416

    work page doi:10.1103/physreva.83.063416 2011

  19. [19]

    O. E. Alon, A. I. Streltsov, and L. S. Cederbaum, The Journal of Chemical Physics127, 154103 (2007), ISSN 0021-9606

    work page 2007

  20. [20]

    A. N. Rubtsov, Phys. Rev. E97, 052120 (2018), URL https://link.aps.org/doi/10.1103/PhysRevE.97. 052120

    work page doi:10.1103/physreve.97 2018

  21. [21]

    A. N. Rubtsov, E. A. Stepanov, and A. I. Lichtenstein, Physical Review B102, 224423 (2020)

    work page 2020

  22. [22]

    Y. S. Lyakhova, E. A. Stepanov, and A. N. Rubtsov, Physical Review B105, 035118 (2022)

    work page 2022

  23. [23]

    Y. S. Lyakhova, S. D. Semenov, A. I. Lichten- stein, and A. N. Rubtsov, Phys. Rev. B110, 245134 (2024), URLhttps://link.aps.org/doi/10. 1103/PhysRevB.110.245134

    work page 2024

  24. [24]

    Schauerte and P

    T. Schauerte and P. G. J. van Dongen, Phys. Rev. B65, 081105 (2002)

    work page 2002

  25. [25]

    P. W. Leung, Z. Liu, E. Manousakis, M. A. Novotny, and P. E. Oppenheimer, Phys. Rev. B 46, 11779 (1992), URLhttps://link.aps.org/doi/10. 1103/PhysRevB.46.11779

    work page 1992

  26. [26]

    Parola, S

    A. Parola, S. Sorella, S. Baroni, M. Par- rinello, and E. Tosatti, International Jour- nal of Modern Physics B03, 1865 (1989), https://doi.org/10.1142/S0217979289001202, URL https://doi.org/10.1142/S0217979289001202

    work page doi:10.1142/s0217979289001202 1989