Pith. sign in

REVIEW 4 major objections 7 minor 86 references

Enhanced coarsening of charge density waves induced by electron correlation: Machine-learning enabled large-scale dynamical simulations

T0 review · 4 major / 7 minor · reviewed 2026-08-10 · deepseek-v4-flash

Pith's one-line read Electron repulsion speeds up charge-density-wave coarsening in a Hubbard-Holstein model.

desk verdict Hubbard U accelerates CDW coarsening in the adiabatic Hubbard-Holstein model—a new result that holds within the Gutzwiller approximation, with the GA itself as the main soft spot. read the letter →

arxiv 2412.21072 v1 pith:KRNI3U7C submitted 2024-12-30 cond-mat.str-el cond-mat.stat-mechcs.LG

classification cond-mat.str-elcond-mat.stat-mechcs.LG
keywords chargedensitywavecoarseningdynamicsHubbard-HolsteinmodelmachinelearningforcefieldGutzwillerapproximationdisorderscreeningLangevinphaseorderingkinetics
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 tries to establish that in the adiabatic half-filled square-lattice Hubbard-Holstein model, increasing the Hubbard repulsion U accelerates rather than suppresses the coarsening of checkerboard charge-density-wave domains after a thermal quench. It further claims this acceleration is a disorder-screening effect: electron correlations renormalize the effective random on-site potential produced by the Holstein phonons, weakening the pinning that normally slows domain growth. To test this in large systems, the authors build a machine-learned force field trained on Gutzwiller-approximation forces, achieving linear-scaling Langevin dynamics on 200x200 lattices. If true, the result shows that standard universal coarsening classes can be reached in correlated electron systems once correlation-induced screening is accounted for, and it validates a route to multiscale dynamical simulations.

What carries the argument

The carrying mechanism is a machine-learned force field that predicts the local electronic force on each lattice site from a symmetry-invariant description of its neighborhood, trained on forces computed by the Gutzwiller/slave-boson approximation. The local descriptors are built from irreducible representations of the D4 point group, and a deep neural network approximates the force as a function of those descriptors, giving O(1) per-site cost and linear overall scaling. The physical explanation of the enhanced coarsening rests on a self-energy renormalization formula epsilon'_i = epsilon_i + Sigma_i(0), which shows that the variance of the effective disorder potential is reduced by a factor 1/(1+U chi_ii)^2, where chi_ii is the local compressibility.

What would settle it

Run the same thermal-quench coarsening protocol on small systems using an unbiased many-body method such as determinant quantum Monte Carlo or exact diagonalization for U/tnn = 0.6 and 0.9, measure L(t), and check whether the domain growth is still faster than at U/tnn = 0.3 and approaches t^(1/2); if the enhancement disappears or the exponent differs, the reported acceleration is an artifact of the Gutzwiller force training.

Watch

Extended reading notes

Core claim

The central claim is that electron-electron repulsion enhances the growth of charge-density-wave domains in the adiabatic Hubbard-Holstein model, opposite to the naive expectation that Hubbard U suppresses Holstein-driven charge order. After a quench to low temperature, the characteristic domain length L(t) grows faster for larger U, and for U/tnn >= 0.9 it recovers the Allen-Cahn law L(t) ~ t^(1/2), while at small U growth is slow and nearly arrested. The late-stage correlation functions collapse onto a single curve when rescaled by L(t), confirming dynamic scaling invariance. The paper attributes the enhancement to a screening mechanism: the Hubbard interaction reduces the effective random on-site potential generated by the Holstein phonons, analogous to disorder screening in correlated systems, and this weaker pinning allows domains to grow more freely.

Load-bearing premise

The central physical result rests on the Gutzwiller approximation being accurate for the disordered, out-of-equilibrium lattice configurations sampled during domain growth, since the machine-learned forces are trained to reproduce Gutzwiller forces.

Editorial extensions

If this is right

  • For Hubbard U/tnn of at least 0.9, CDW domain growth obeys the Allen-Cahn t^(1/2) law, meaning electron correlations restore the standard curvature-driven coarsening universality class.
  • The late-stage coarsening satisfies dynamic scaling invariance: equal-time correlation functions collapse onto a single time-independent curve when lengths are rescaled by L(t).
  • The Hubbard interaction screens the Holstein interaction in the adiabatic limit, effectively reducing the random on-site potential that pins CDW domains and thereby accelerating their growth.
  • Machine-learned force fields with locality-based descriptors provide a linear-scaling algorithm suitable for large-scale dynamical simulations of correlated electron systems, going beyond the limits of direct many-body methods.
  • This is presented as the first systematic study of electron-correlation effects on phase-ordering dynamics, opening a new direction in non-equilibrium physics of emergent orders.

Reading between the lines

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

  • If the screening mechanism generalizes beyond the specific model, moderately correlated materials could show faster CDW ordering kinetics than simple electron-phonon models predict, a difference that may be observable in ultrafast pump-probe experiments.
  • The same machine-learning pipeline could be applied to other emergent orders such as spin density waves or orbital order, where the feedback between quasiparticles and the order parameter may produce different coarsening universality classes.
  • A direct test of the approximation dependence would be to train the identical ML architecture on forces from exact diagonalization or determinant quantum Monte Carlo on small systems; if those forces reproduce the same enhancement, the result is not an artifact of the Gutzwiller approximation.
  • The paper's mechanism implies that the coarsening rate is controlled by the local compressibility chi_ii, so tuning U or the filling to maximize chi_ii could systematically accelerate or decelerate domain growth in experiments.
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 / 7 minor

Summary. This paper presents a machine-learning force-field approach to simulate the coarsening dynamics of charge density waves in the adiabatic square-lattice Hubbard-Holstein model. The force on each lattice displacement is computed in the Gutzwiller approximation and learned by a six-layer neural network acting on D4-symmetric local descriptors, enabling 200x200 Langevin simulations. The authors report that increasing the Hubbard repulsion U accelerates CDW domain growth, with L(t) recovering the Allen-Cahn t^(1/2) law for U/tnn >= 0.9, and attribute this to screening of the Holstein potential by electron correlations. They also report dynamic scaling collapse of the correlation function at late times.

Significance. If the conclusions hold, the paper provides a novel and nontrivial result: electron repulsion, which competes with CDW order, accelerates rather than suppresses domain coarsening, and the proposed ML protocol demonstrates a promising route to linear-scaling simulations of correlated electron dynamics. The surrogate itself is carefully validated against the Gutzwiller reference (force standard deviations about 0.006 in the units used, and correlation-function benchmarks on 14x14 systems), which gives confidence in the ML step. The main limitation is that the reference method (Gutzwiller approximation) is not benchmarked against an exact many-body method for the non-equilibrium domain-wall states relevant to coarsening; the physical claim is therefore conditional on the accuracy of GA in this regime. The paper also leaves unspecified several simulation parameters (e.g., damping, cutoff radius) needed to reproduce and assess the results.

major comments (4)
  1. [Sec. II, ML force field and Fig. 2] The benchmarks in Fig. 1 and Fig. 2 only establish that the ML model reproduces the Gutzwiller-approximation (GA) forces, not that GA is accurate for the domain-wall-rich configurations sampled during coarsening. Since the central physical claim is about the Hubbard-Holstein model, the approximation dependence is load-bearing. Please provide a direct benchmark of GA against an exact/near-exact method (e.g., determinant quantum Monte Carlo or exact diagonalization on small lattices) for representative static lattice configurations with domain walls, and state the expected accuracy of GA for this regime.
  2. [Sec. II, Eq. (3) and locality paragraph] The cutoff radius Rc in Eq. (3) is never specified and no convergence test with respect to Rc is reported. The linear-scaling claim relies on this truncation; without a test that the learned forces (or the resulting L(t)) have converged with Rc, the surrogate may be missing longer-range correlation effects. Please report Rc, the corresponding number of neighbors, and a convergence study.
  3. [Sec. II, Eq. (2)] The Langevin equation (2) contains the damping gamma and mass m, but neither is specified in the text, and the integration timestep and total simulation time are not given. The coarsening rate and the exponent in Fig. 4a depend on the dynamics being in the overdamped regime and on the thermostat implementation. Please provide gamma, m, the numerical integrator, and the dimensionless time convention; otherwise the quantitative Allen-Cahn result is not reproducible.
  4. [Sec. III, Eqs. (7)-(8)] The disorder-screening mechanism is inferred from a DMFT result for static random disorder, Eq. (8), but the Holstein displacement Q_i is a dynamical degree of freedom and the simulations do not directly measure the screening of the potential or the domain-wall barrier. Please either test this mechanism directly (for example, by comparing the GA potential barrier for a domain wall as a function of U, or by computing the variance of the effective on-site potential along the trajectory) or clearly label it as a plausible but unverified interpretation.
minor comments (7)
  1. [Fig. 4] The scaling collapse is shown for U/tnn = 0.6, 0.9, 1.2, 1.5 but not for U/tnn = 0.3; the text's statement that 'the late-stage coarsening dynamics ... obeys dynamic scaling invariance' is broader than the presented evidence. Please show the U=0.3 collapse (or explain its absence) and qualify the claim.
  2. [Eq. (4)] Please clarify whether n_i is the Gutzwiller expectation value obtained simultaneously with the force computation, and how the phase exp(iQ·r_i) is treated for finite systems with periodic boundary conditions.
  3. [Methods, 'Neural network and training'] The training dataset description is incomplete. Please specify the system size and the range of lattice configurations used to generate the 300 snapshots, the number of training/validation splits, and the number of independent ML models trained for each U.
  4. [Throughout] There are several typographical errors: 'U/nn = 0.3' should be 'U/tnn'; 'Gutswiller' in the Fig. 2 caption should be 'Gutzwiller'; 'indicting' in Sec. II should be 'indicating'; and 'two eigenvalues problems' in Methods should be 'two eigenvalue problems'.
  5. [Methods, 'Gutzwiller approximation'] The notation rho_{ij,sigma} in Eq. (11) and Delta_{i,sigma} in Eq. (13) should be explicitly defined with indices, and the sentence about the two eigenvalue problems should be rewritten for clarity.
  6. [Reference [83]] Reference [83] is cited as 'unpublished'; it is used both for the slow-coarsening U=0.3 behavior and for the t^(1/2) weak-coupling limit. Please clarify whether this reference is under review or replace it with a publicly available version.
  7. [Fig. 4a] The data points for L(t) are shown as symbols without error bars. Since the correlation function is averaged over 70 runs, please indicate the statistical uncertainty (e.g., standard error of the mean) for L(t).

Circularity Check

0 steps flagged · score 2.0 of 10

No constructional circularity: the coarsening enhancement is a simulated output of GA-based Langevin dynamics, not an input; only a minor self-citation to an unpublished companion paper is used as a qualitative comparison limit.

full rationale

The derivation chain is self-contained: Eq. (1) defines the Hubbard-Holstein model, Eq. (2) defines Langevin dynamics, Eqs. (9)-(13) define the Gutzwiller force calculation, and the ML model is trained with loss Eq. (14) to reproduce the GA forces. The characteristic length L(t) in Fig. 4 is measured from the time-dependent correlation function C(r,t), not fitted to the Allen-Cahn law; the growth law and the U-dependent enhancement emerge from the dynamics. The force field is a surrogate for GA, but the benchmark in Figs. 1-2 checks the surrogate against GA rather than injecting the coarsening result into the training, so there is no fitted-input-called-prediction step. The use of GA is an approximation-validity risk, not a circularity, because the paper does not claim an exact many-body benchmark for the non-equilibrium domain-wall configurations. The mechanism section invokes the published DMFT disorder-screening result (Ref. 84) and a screening term visible in the GA Hamiltonian, which is an interpretive analogy rather than a definitional identity. The only mild self-citation concern is Ref. 83, the authors' unpublished companion paper, which is cited as the weak-coupling Holstein coarsening baseline used to interpret the U >= 0.9 branch. That baseline is not independently checkable, but it is used as a qualitative comparison endpoint, not as an input to the equations that produce L(t); therefore it does not make the central claim circular. Overall score 2 reflects this minor non-load-bearing self-citation, with no significant circularity.

Assumptions & free parameters 3 free parameters · 6 assumptions · 0 invented entities

The physical conclusion rests mainly on the Gutzwiller approximation, the adiabatic/semiclassical treatment of phonons, the nearsightedness assumption underlying the force-field truncation, and the transfer of the DMFT disorder-screening formula (Eq 8) to the HH model. No new particles, forces, or dimensions are introduced.

free parameters (3)
  • Neural network weights of the ML force field = not disclosed (trained on 300 GA snapshots per U)
    The central claim depends on the surrogate reproducing GA forces; the weights are fit to data, but they are not physical constants and do not enter the screening mechanism.
  • Descriptor reference basis f^Gamma_ref = weighted average within local environment
    Introduced by hand for symmetry-invariant descriptors; its specific construction could affect force accuracy but is not a physical parameter.
  • Cutoff radius Rc = not stated
    Determines the size of the local environment; is a hand-chosen hyperparameter of the ML model, and its value is not given in the text.
assumptions (6)
  • domain assumption Adiabatic (Born-Oppenheimer) approximation for the lattice dynamics is valid.
    Invoked in Section II after Eq. 2: 'the evolution of the CDW state can be effectively described by using the adiabatic approximation.' The conclusion depends on the quasiparticle field relaxing faster than the order parameter.
  • domain assumption Semiclassical treatment of phonons as classical variables is valid for CDW physics.
    Stated in Section II: the authors rely on agreement between hybrid Monte Carlo and determinant QMC for equilibrium CDW phases to justify classical phonons in the dynamics.
  • domain assumption The Gutzwiller approximation accurately captures the relevant electron correlation effects.
    Stated in Section II: 'The Gutzwiller approximation (GA) also accurately captures correlation effects such as bandwidth renormalization and disorder screening.' All simulated forces and the coarsening results inherit this approximation.
  • domain assumption Nearsightedness principle holds so that local forces depend only on a fixed neighborhood with cutoff Rc.
    Stated in the Introduction and Section II: 'the locality principle indicates that local forces that drive domain growth only depend on order-parameter configurations in the corresponding immediate surroundings.' The linear-scaling ML architecture is built on this.
  • domain assumption The DMFT disorder-screening formula, Eqs 7-8 from Ref 84, transfers to the adiabatic Hubbard-Holstein model with Holstein coupling treated as a random on-site potential.
    Used in Section III to explain the enhanced coarsening: Eq 8 is presented as a known result and applied without derivation to the HH model.
  • standard math Universal approximation theorem guarantees the neural network can represent the force function.
    Invoked in the Introduction: 'a deep-learning neural network can then be trained to accurately capture this complex dependence,' relying on standard approximation theorems.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Enhanced coarsening of charge density waves induced by electron correlation: Machine-learning enabled large-scale dynamical simulations." pith.science (2026). https://pith.science/paper/KRNI3U7C

@misc{pith2026241221072,
  author       = {Pith},
  title        = {Pith review of: Enhanced coarsening of charge density waves induced by electron correlation: Machine-learning enabled large-scale dynamical simulations},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/KRNI3U7C}},
  note         = {Machine review of arXiv:2412.21072}
}
read the original abstract

The phase ordering kinetics of emergent orders in correlated electron systems is a fundamental topic in non-equilibrium physics, yet it remains largely unexplored. The intricate interplay between quasiparticles and emergent order-parameter fields could lead to unusual coarsening dynamics that is beyond the standard theories. However, accurate treatment of both quasiparticles and collective degrees of freedom is a multi-scale challenge in dynamical simulations of correlated electrons. Here we leverage modern machine learning (ML) methods to achieve a linear-scaling algorithm for simulating the coarsening of charge density waves (CDWs), one of the fundamental symmetry breaking phases in functional electron materials. We demonstrate our approach on the square-lattice Hubbard-Holstein model and uncover an intriguing enhancement of CDW coarsening which is related to the screening of on-site potential by electron-electron interactions. Our study provides fresh insights into the role of electron correlations in non-equilibrium dynamics and underscores the promise of ML force-field approaches for advancing multi-scale dynamical modeling of correlated electron systems.

Figures

Figures reproduced from arXiv: 2412.21072 by the authors.

Figure 1
Figure 1. FIG. 1 [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2. Comparison of lattice correlation function [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. FIG. 3. Snapshots of local CDW parameter [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: a. At U/nn = 0.3, L(t) grows very slowly and stops at the late stage, similar to the slow coarsening dynam￾ics observed in the strong coupling limit of the Holstein model [14]. As the Hubbard U increases, the growth of the characteristic domain length becomes faster an…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

86 extracted references · 48 canonical work pages

  1. [83]

    Y. Yang, S. Zhang, H. Jang, and G.-W. Chern, Anoma- lous coarsening of charge density wave order in holstein model, unpublished

  2. [1]

    Bray, Theory of phase-ordering kinetics, Advances in Physics 43, 357 (1994)

    A. Bray, Theory of phase-ordering kinetics, Advances in Physics 43, 357 (1994)

  3. [2]

    Onuki, Phase Transition Dynamics(Cambridge Uni- versity Press, 2002)

    A. Onuki, Phase Transition Dynamics(Cambridge Uni- versity Press, 2002)

  4. [3]

    V. W. Sanjay Puri, Kinetics of phase transitions(CRC Press, 2009)

  5. [4]

    L. F. Cugliandolo, Coarsening phenomena, Comptes Rendus. Physique 16, 257 (2015)

  6. [5]

    I. M. Lifshitz and V. V. Slyozov, The kinetics of pre- cipitation from supersaturated solid solutions, Journal of physics and chemistry of solids 19, 35 (1961)

  7. [6]

    Lifshitz, Kinetics of ordering during second-order phase 9 transitions, Sov

    I. Lifshitz, Kinetics of ordering during second-order phase 9 transitions, Sov. Phys. JETP 15, 939 (1962)

  8. [7]

    Wagner, Theory of precipitate change by redissolution, Z

    C. Wagner, Theory of precipitate change by redissolution, Z. Elektrochem 65, 581 (1961)

Show all 86 references
  1. [8]

    T. Wu, H. Mayaffre, S. Kr¨ amer, M. Horvati´ c, C. Berthier, R. L. W. N. Hardy, D. A. Bonn, and M.-H. Julien, Magnetic-field-induced charge-stripe order in the high- temperature superconductor yba2cu3oy, Nature 477, 191 (2011)

  2. [9]

    Hirata, Y

    T. Hirata, Y. Yamakawa, S. Onari, and H. Kontani, Un- conventional orbital charge density wave mechanism in the transition metal dichalcogenide 1 t − tas2, Phys. Rev. Res. 3, L032053 (2021)

  3. [10]

    J. G. Park, B. A. Collins, L. E. Darago, T. Runˇ cevski, M. E. Ziebel, M. L. Aubrey, H. Z. Jiang, E. Velasquez, M. A. Green, J. D. Goodpaster, et al., Magnetic order- ing through itinerant ferromagnetism in a metal–organic framework, Nature Chemistry 13, 594 (2021)

  4. [11]

    Y. Hu, T. Zhang, C. C. Dongming Zhao, S. Ding, W. Yang, X. Wang, C. Li, H. Wang, D. Feng, and T. Zhang, Real-space observation of incommensurate spin density wave and coexisting charge density wave on cr (001) surface, Nat Commun 13, 445 (2022)

  5. [12]

    P. C. Hohenberg and B. I. Halperin, Theory of dynamic critical phenomena, Rev. Mod. Phys. 49, 435 (1977)

  6. [13]

    Belitz, T

    D. Belitz, T. R. Kirkpatrick, and R. Saha, Phase-ordering dynamics in itinerant quantum ferromagnets, Phys. Rev. B 75, 144418 (2007)

  7. [15]

    Ghosh, S

    S. Ghosh, S. Zhang, C. Cheng, and G.-W. Chern, Kinet- ics of orbital ordering in cooperative jahn-teller models: Machine-learning enabled large-scale simulations, Phys. Rev. Mater. 8, 123602 (2024)

  8. [16]

    Y. Fan, S. Zhang, and G.-W. Chern, Coarsening of chi- ral domains in itinerant electron magnets: A machine learning force-field approach, Phys. Rev. B 110, 245105 (2024)

  9. [17]

    Kohn, Density functional and density matrix method scaling linearly with the number of atoms, Phys

    W. Kohn, Density functional and density matrix method scaling linearly with the number of atoms, Phys. Rev. Lett. 76, 3168 (1996)

  10. [18]

    Prodan and W

    E. Prodan and W. Kohn, Nearsightedness of electronic matter, Proc. Natl. Acad. Sci. USA 102, 11635 (2005)

  11. [19]

    Silver and H

    R. Silver and H. R¨ oder, Densities of states of mega- dimensional hamiltonian matrices, International Journal of Modern Physics C 05, 735 (1994)

  12. [20]

    Weiße, G

    A. Weiße, G. Wellein, A. Alvermann, and H. Fehske, The kernel polynomial method, Rev. Mod. Phys. 78, 275 (2006)

  13. [21]

    Wang, G.-W

    Z. Wang, G.-W. Chern, C. D. Batista, and K. Bar- ros, Gradient-based stochastic estimation of the density matrix, The Journal of Chemical Physics 148, 094107 (2018)

  14. [22]

    Behler and M

    J. Behler and M. Parrinello, Generalized neural-network representation of high-dimensional potential-energy sur- faces, Phys. Rev. Lett. 98, 146401 (2007)

  15. [23]

    A. P. Bart´ ok, M. C. Payne, R. Kondor, and G. Cs´ anyi, Gaussian approximation potentials: The accuracy of quantum mechanics, without the electrons, Phys. Rev. Lett. 104, 136403 (2010)

  16. [24]

    Z. Li, J. R. Kermode, and A. De Vita, Molecular dy- namics with on-the-fly machine learning of quantum- mechanical forces, Phys. Rev. Lett. 114, 096405 (2015)

  17. [25]

    A. V. Shapeev, Moment tensor potentials: A class of sys- tematically improvable interatomic potentials, Multiscale Modeling & Simulation 14, 1153 (2016)

  18. [26]

    V. Botu, R. Batra, J. Chapman, and R. Ramprasad, Ma- chine learning force fields: Construction, validation, and outlook, The Journal of Physical Chemistry C 121, 511 (2017)

  19. [27]

    J. S. Smith, O. Isayev, and A. E. Roitberg, Ani-1: an extensible neural network potential with dft accuracy at force field computational cost, Chem. Sci. 8, 3192 (2017)

  20. [28]

    Zhang, J

    L. Zhang, J. Han, H. Wang, R. Car, and W. E, Deep potential molecular dynamics: A scalable model with the accuracy of quantum mechanics, Phys. Rev. Lett. 120, 143001 (2018)

  21. [29]

    Behler, Perspective: Machine learning potentials for atomistic simulations, The Journal of Chemical Physics 145, 170901 (2016)

    J. Behler, Perspective: Machine learning potentials for atomistic simulations, The Journal of Chemical Physics 145, 170901 (2016)

  22. [30]

    V. L. Deringer, M. A. Caro, and G. Csanyi, Machine learning interatomic potentials as emerging tools for ma- terials science, Advanced Materials 31, 1902765 (2019)

  23. [31]

    R. T. McGibbon, A. G. Taube, A. G. Donchev, K. Siva, F. Hernandez, C. Hargus, K.-H. Law, J. L. Klepeis, and D. E. Shaw, Improving the accuracy of m¨ oller-plesset per- turbation theory with neural networks, The Journal of Chemical Physics 147, 161725 (2017)

  24. [32]

    H. Suwa, J. S. Smith, N. Lubbers, C. D. Batista, G.-W. Chern, and K. Barros, Machine learning for molecular dynamics with strongly correlated electrons, Phys. Rev. B 99, 161107 (2019)

  25. [33]

    Chmiela, A

    S. Chmiela, A. Tkatchenko, H. E. Sauceda, I. Poltavsky, K. T. Sch¨ utt, and K.-R. M¨ uller, Machine learning of ac- curate energy-conserving molecular force fields, Science Advances 3, e1603015 (2017)

  26. [34]

    Chmiela, H

    S. Chmiela, H. E. Sauceda, K.-R. M¨ uller, and A. Tkatchenko, Towards exact molecular dynamics sim- ulations with machine-learned force fields, Nature Com- munications 9, 3887 (2018)

  27. [35]

    H. E. Sauceda, M. Gastegger, S. Chmiela, K.-R. M¨ uller, and A. Tkatchenko, Molecular force fields with gradient- domain machine learning (GDML): Comparison and syn- ergies with classical force fields, The Journal of Chemical Physics 153, 124109 (2020)

  28. [36]

    Cybenko, Approximation by superpositions of a sig- moidal function, Math

    G. Cybenko, Approximation by superpositions of a sig- moidal function, Math. Control Signal Systems 2, 303 (1989)

  29. [37]

    Hornik, M

    K. Hornik, M. Stinchcombe, and H. White, Multilayer feedforward networks are universal approximators, Neu- ral Networks 2, 359 (1989)

  30. [38]

    Barron, Universal approximation bounds for super- positions of a sigmoidal function, IEEE Transactions on Information Theory 39, 930 (1993)

    A. Barron, Universal approximation bounds for super- positions of a sigmoidal function, IEEE Transactions on Information Theory 39, 930 (1993)

  31. [39]

    Gr¨ uner, The dynamics of charge-density waves, Rev

    G. Gr¨ uner, The dynamics of charge-density waves, Rev. Mod. Phys. 60, 1129 (1988)

  32. [40]

    Gr¨ uner,Density Waves In Solids, 1st ed

    G. Gr¨ uner,Density Waves In Solids, 1st ed. (CRC Press, 1994)

  33. [41]

    X. Shi, W. You, Y. Zhang, Z. Tao, P. M. Oppe- neer, X. Wu, R. Thomale, K. Rossnagel, M. Bauer, H. Kapteyn, and M. Murnane, Ultrafast electron calorimetry uncovers a new long-lived metastable state in 1 t-tase2 mediated by mode-selective electron-phonon coupling, Sci. Adv. 5, 4...

  34. [42]

    Maklar, Y

    J. Maklar, Y. W. Windsor, C. W. Nicholson, M. Pup- pin, P. Walmsley, V. Esposito, M. Porer, J. Rittmann, D. Leuenberger, M. Kubli, M. Savoini, E. Abreu, S. L. Johnson, P. Beaud, G. Ingold, U. Staub, I. R. Fisher, 10 R. Ernstorfer, M. Wolf, and L. Rettig, Nonequilibrium charge-...

  35. [43]

    S. Duan, W. Xia, C. Huang, S. Wang, L. Gu, H. Liu, D. Xiang, D. Qian, Y. Guo, and W. Zhang, Ultra- fast switching from the charge density wave phase to a metastable metallic state in 1 t−tise2, Phys. Rev. Lett. 130, 226501 (2023)

  36. [44]

    Boschini, M

    F. Boschini, M. Zonno, and A. Damascelli, Time-resolved arpes studies of quantum materials, Rev. Mod. Phys. 96, 015003 (2024)

  37. [45]

    Berger, P

    E. Berger, P. Val´ aˇ sek, and W. von der Linden, Two- dimensional hubbard-holstein model, Phys. Rev. B 52, 4806 (1995)

  38. [46]

    Bauer, Competing interactions and symmetry breaking in the hubbard-holstein model, Europhysics Letters 90, 27002 (2010)

    J. Bauer, Competing interactions and symmetry breaking in the hubbard-holstein model, Europhysics Letters 90, 27002 (2010)

  39. [47]

    Bauer and A

    J. Bauer and A. C. Hewson, Competition between anti- ferromagnetic and charge order in the hubbard-holstein model, Phys. Rev. B 81, 235113 (2010)

  40. [48]

    Fehske, G

    H. Fehske, G. Wellein, G. Hager, A. Weiße, and A. R. Bishop, Quantum lattice dynamical effects on single- particle excitations in one-dimensional mott and peierls insulators, Phys. Rev. B 69, 165115 (2004)

  41. [49]

    Kumar and J

    S. Kumar and J. van den Brink, Charge ordering and magnetism in quarter-filled hubbard-holstein model, Phys. Rev. B 78, 155123 (2008)

  42. [50]

    Weber and M

    M. Weber and M. Hohenadler, Two-dimensional holstein- hubbard model: Critical temperature, ising universality, and bipolaron liquid, Phys. Rev. B 98, 085405 (2018)

  43. [51]

    Holstein, Studies of polaron motion: Part I

    T. Holstein, Studies of polaron motion: Part I. the molecular-crystal model, Annals of Physics 8, 325 (1959)

  44. [52]

    R. M. Noack, D. J. Scalapino, and R. T. Scalettar, Charge-density-wave and pairing susceptibilities in a two- dimensional electron-phonon model, Phys. Rev. Lett. 66, 778 (1991)

  45. [53]

    Zhang, W.-T

    Y.-X. Zhang, W.-T. Chiu, N. C. Costa, G. G. Batrouni, and R. T. Scalettar, Charge order in the holstein model on a honeycomb lattice, Phys. Rev. Lett. 122, 077602 (2019)

  46. [54]

    C. Chen, X. Y. Xu, Z. Y. Meng, and M. Hohenadler, Charge-density-wave transitions of dirac fermions cou- pled to phonons, Phys. Rev. Lett. 122, 077601 (2019)

  47. [55]

    Hohenadler and G

    M. Hohenadler and G. G. Batrouni, Dominant charge density wave correlations in the holstein model on the half-filled square lattice, Phys. Rev. B 100, 165114 (2019)

  48. [56]

    Hubbard, Electron correlations in narrow energy bands, Proc

    J. Hubbard, Electron correlations in narrow energy bands, Proc. R. Soc. Lond. A 276, 238 (1963)

  49. [57]

    R. T. Clay and R. P. Hardikar, Intermediate phase of the one dimensional half-filled hubbard-holstein model, Phys. Rev. Lett. 95, 096401 (2005)

  50. [58]

    E. A. Nowadnick, S. Johnston, B. Moritz, R. T. Scalet- tar, and T. P. Devereaux, Competition between antiferro- magnetic and charge-density-wave order in the half-filled hubbard-holstein model, Phys. Rev. Lett. 109, 246404 (2012)

  51. [59]

    Johnston, E

    S. Johnston, E. A. Nowadnick, Y. F. Kung, B. Moritz, R. T. Scalettar, and T. P. Devereaux, Determinant quan- tum monte carlo study of the two-dimensional single- band hubbard-holstein model, Phys. Rev. B 87, 235133 (2013)

  52. [60]

    E. A. Nowadnick, S. Johnston, B. Moritz, and T. P. Dev- ereaux, Renormalization of spectra by phase competition in the half-filled hubbard-holstein model, Phys. Rev. B 91, 165127 (2015)

  53. [61]

    Costa, K

    N. Costa, K. Seki, S. Yunoki, and S. Sorella, Phase di- agram of the two-dimensional hubbard-holstein model, Commun Phys 3, 80 (2020)

  54. [62]

    Esterlis, S

    I. Esterlis, S. A. Kivelson, and D. J. Scalapino, Pseudo- gap crossover in the electron-phonon system, Phys. Rev. B 99, 174516 (2019)

  55. [63]

    Marx and J

    D. Marx and J. Hutter, Ab Initio Molecular Dynamics: Basic Theory and Advanced Methods(Cambridge Uni- versity Press, 2009)

  56. [64]

    Hellmann, T

    S. Hellmann, T. Rohwer, M. Kall¨ ane, K. Hanff, C. Sohrt, A. Stange, A. Carr, M. M. Murnane, H. C. Kapteyn, L. Kipp, M. Bauer, and K. Rossnagel, Time-domain clas- sification of charge-density-wave insulators, Nat Com- mun 3, 1069 (2012)

  57. [65]

    C. J. Sayers, H. Hedayat, A. Ceraso, F. Museur, M. Cat- telan, L. S. Hart, L. S. Farrar, S. Dal Conte, G. Cerullo, C. Dallera, E. Da Como, and E. Carpene, Coherent phonons and the interplay between charge density wave and mott phases in 1t − Tase2, Phys. Rev. B 102, 161105 (2020)

  58. [66]

    S.-Q. Hu, H. Zhao, C. Lian, X.-B. Liu, M.-X. Guan, and S. Meng, Tracking photocarrier-enhanced electron- phonon coupling in nonequilibrium, npj Quantum Mater. 7, 14 (2022)

  59. [67]

    Zhang, S

    P. Zhang, S. Zhang, and G.-W. Chern, Descriptors for machine learning model of generalized force field in con- densed matter systems (2022), arXiv:2201.00798 [cond- mat.str-el]

  60. [68]

    Zhang, P

    S. Zhang, P. Zhang, and G.-W. Chern, Anomalous phase separation in a correlated electron system: Machine- learning enabled large-scale kinetic monte carlo simula- tions, Proceedings of the National Academy of Sciences 119, e2119957119 (2022)

  61. [69]

    Cheng, S

    C. Cheng, S. Zhang, and G.-W. Chern, Machine learn- ing for phase ordering dynamics of charge density waves, Phys. Rev. B 108, 014301 (2023)

  62. [70]

    Zhang, P

    P. Zhang, P. Saha, and G.-W. Chern, Machine learning dynamics of phase separation in correlated electron mag- nets (2020), arXiv:2006.04205 [cond-mat.str-el]

  63. [72]

    Zhang and G.-W

    P. Zhang and G.-W. Chern, Machine learning nonequilib- rium electron forces for spin dynamics of itinerant mag- nets, npj Computational Materials 9, 32 (2023)

  64. [73]

    Cheng, S

    X. Cheng, S. Zhang, P. C. H. Nguyen, S. Azarfar, G.-W. Chern, and S. S. Baek, Convolutional neural networks for large-scale dynamical modeling of itinerant magnets, Phys. Rev. Res. 5, 033188 (2023)

  65. [74]

    M. C. Gutzwiller, Effect of correlation on the ferromag- netism of transition metals, Phys. Rev. Lett. 10, 159 (1963)

  66. [75]

    M. C. Gutzwiller, Effect of correlation on the ferromag- netism of transition metals, Phys. Rev.134, A923 (1964)

  67. [76]

    M. C. Gutzwiller, Correlation of electrons in a narrow s band, Phys. Rev. 137, A1726 (1965)

  68. [77]

    B¨ unemann, F

    J. B¨ unemann, F. Gebhard, and W. Weber, Gutzwiller- correlated wave functions for degenerate bands: exact results in infinite dimensions, Journal of Physics: Con- densed Matter 9, 7343 (1997)

  69. [78]

    B¨ unemann, W

    J. B¨ unemann, W. Weber, and F. Gebhard, Multiband 11 gutzwiller wave functions for general on-site interactions, Phys. Rev. B 57, 6896 (1998)

  70. [79]

    B¨ unemann, F

    J. B¨ unemann, F. Gebhard, and R. Thul, Landau- gutzwiller quasiparticles, Phys. Rev. B 67, 075103 (2003)

  71. [80]

    Fabrizio, Gutzwiller description of non-magnetic mott insulators: Dimer lattice model, Phys

    M. Fabrizio, Gutzwiller description of non-magnetic mott insulators: Dimer lattice model, Phys. Rev. B 76, 165110 (2007)

  72. [81]

    Lanat` a, P

    N. Lanat` a, P. Barone, and M. Fabrizio, Fermi-surface evolution across the magnetic phase transition in the kondo lattice model, Phys. Rev. B 78, 155127 (2008)

  73. [82]

    Lanat` a, H

    N. Lanat` a, H. U. R. Strand, X. Dai, and B. Hells- ing, Efficient implementation of the gutzwiller variational method, Phys. Rev. B 85, 035133 (2012)

  74. [84]

    Tanaskovi´ c, V

    D. Tanaskovi´ c, V. Dobrosavljevi´ c, E. Abrahams, and G. Kotliar, Disorder screening in strongly correlated sys- tems, Phys. Rev. Lett. 91, 066603 (2003)

  75. [85]

    Kondor, A novel set of rotationally and transla- tionally invariant features for images based on the non-commutative bispectrum (2007), arXiv:cs/0701127 [cs.CV]

    R. Kondor, A novel set of rotationally and transla- tionally invariant features for images based on the non-commutative bispectrum (2007), arXiv:cs/0701127 [cs.CV]

  76. [86]

    Zhang and G.-W

    P. Zhang and G.-W. Chern, Arrested phase separation in double-exchange models: Large-scale simulation en- abled by machine learning, Phys. Rev. Lett. 127, 146401 (2021)

  77. [87]

    Paszke, S

    A. Paszke, S. Gross, F. Massa, A. Lerer, J. Bradbury, G. Chanan, T. Killeen, Z. Lin, N. Gimelshein, L. Antiga, et al., Pytorch: An imperative style, high-performance deep learning library, Advances in neural informa- tion processing systems 32, 10.48550/arXiv.1912.01703 (2019)

  78. [88]

    D. P. Kingma and J. Ba, Adam: A method for stochastic optimization (2017), arXiv:1412.6980 [cs.LG]

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.