Pith. sign in

REVIEW 3 major objections 5 minor 1 cited by

Strong coupling impurity solver based on quantics tensor cross interpolation

T0 review · 3 major / 5 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read Quantics tensor cross interpolation turns the one-crossing approximation into a practical nonequilibrium impurity solver, reproducing reference spectra to about four digits while cutting DMFT cost by roughly two orders of magnitude.

desk verdict A credible QTCI-based OCA solver with real benchmarks, but the low-eta metallic accuracy claim rests on an untested tmax-truncation assumption that a referee should pin down. read the letter →

arxiv 2411.19026 v2 pith:WYTVQM5X submitted 2024-11-28 cond-mat.str-el

Aaram J. Kim , Philipp Werner This is my paper · ORCID
classification cond-mat.str-el
keywords quanticstensorcrossinterpolationone-crossingapproximationnonequilibriumdynamicalmean-fieldtheoryAndersonimpuritymodelHubbardpseudoparticleformalismsteadystatetrain
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 develops a strong-coupling impurity solver for the one-crossing approximation (OCA) that evaluates the multi-dimensional integrals with quantics tensor cross interpolation (QTCI), and demonstrates it on the Anderson impurity model and on steady-state dynamical mean-field theory (DMFT) for the Hubbard model. The central claim is that the QTCI factorisation reproduces the OCA pseudoparticle self-energy and the physical Green's functions to about four digits of accuracy, while cutting the numerical cost of the presented DMFT calculations by roughly two orders of magnitude compared with conventional quadrature. This matters because higher-order self-consistent hybridisation expansions beyond the noncrossing approximation were previously too costly for routine nonequilibrium DMFT, and QTCI removes that bottleneck. The paper also shows that the fused-leg tensor representation can fit the discontinuous lesser-component integrand accurately, provided the initial pivots are placed near the discontinuities.

What carries the argument

The central object is the quantics tensor cross interpolation of the OCA integrand $\sigma(v_1,v_2,v_3)$, a matrix-valued weight formed from products of pseudoparticle Green's functions and hybridisation functions on a fused ring-shaped Keldysh contour. Each time-difference variable $v_i$ is written with $R$ binary digits, which maps the function to a tensor; cross interpolation compresses this tensor into a tensor train with bond dimension $N_{\mathrm{bond}}$. In the fused-leg representation, bits of the same scale from the three variables are grouped into tensors $F_j$, which approximately classifies the continuous domains of the discontinuous lesser component. The work this does is to convert the expensive $t^{2n}_{\max}$ integrals of order-$n$ hybridisation expansion into products of one-dimensional integrals, with measurement cost $O(N_\omega R(2n-1)N_{\mathrm{bond}}^3)$.

What would settle it

Evaluate the OCA self-energy of a metallic Anderson impurity model at a small fixed damping $\eta$ with the QTCI solver at $t_{\max}=6000$, then repeat at $t_{\max}=12000$; if the integrated self-energy or spectral function changes by more than the claimed four digits, the truncation assumption is falsified. A complementary check is to compare the QTCI-fitted integrand $\sigma$ against a direct dense-grid evaluation on a parameter set with slowly decaying hybridisation, where the reported four-order-of-magnitude suppression of the fit error should still hold.

Watch

Extended reading notes

Core claim

On the paper's own terms, the discovery is that QTCI can serve as the workhorse of an OCA impurity solver in the steady-state pseudoparticle formalism. Writing the OCA self-energy contribution as a three-dimensional integral over contour-time difference variables $v_1,v_2,v_3$ on a ring-shaped contour, the integrand $\sigma(v_1,v_2,v_3)$ is encoded in a quantics binary representation and factorised into a tensor train; in the fused representation the tensors $F_j$ combine bits of the same scale, and because the Fourier weight $e^{-i\omega \sum_i v_i}$ factorises, the frequency-dependent self-energy becomes a product of one-dimensional matrix contractions. The same strategy evaluates the vertex correction to the physical Green's function. The paper reports that the fit error in $\sigma$ is about four orders of magnitude below the signal, that the self-consistent OCA solutions match conventional L-shaped contour results, and that one DMFT iteration with $2^{18}$ time points takes about 10 minutes on a single processor.

Load-bearing premise

The whole calculation assumes that by the cutoff time $t_{\max}$ the pseudoparticle Green's functions and hybridisation functions have decayed to negligible size, so that truncating the ring-shaped contour at $t_{\max}$ does not change the result.

Editorial extensions

If this is right

  • Equilibrium OCA spectral functions for the half-filled Hubbard model on the Bethe lattice can be computed with the real-time steady-state solver at $t_{\max}\approx 6000$, reaching the conventional reference spectra as $\eta\to 0$.
  • Nonequilibrium steady-state spectra of photodoped Mott insulators can be obtained self-consistently by imposing a nonequilibrium distribution function $f_{\mathrm{neq}}(\omega)$, with the doublon/holon quasi-particle peaks emerging at the imposed chemical potentials.
  • A single DMFT iteration in the presented OCA calculation costs about 10 minutes on one processor for $2^{18}$ time points, which makes parameter scans in steady-state DMFT feasible.
  • The same fused-bit strategy should extend to higher-order self-consistent hybridisation expansions, since the discontinuities of the lesser components at order $n>2$ can in principle be captured by the fused representation.
  • The damping factor $\eta$ genuinely broadens spectra, so accurate low-temperature metallic solutions require $\eta\lesssim 0.001$; this is an intrinsic property of the pseudoparticle steady-state formalism rather than a defect of QTCI.

Reading between the lines

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

  • Because the measurement of the frequency-dependent self-energy, scaling as $O(N_\omega R(2n-1)N_{\mathrm{bond}}^3)$, dominates the cost, a sparse adaptive frequency grid could further reduce the expense without altering the QTCI fit.
  • If the fused-leg representation handles branch discontinuities at general order $n$, the same solver structure should carry over to third-order (TOA) and to multiorbital impurities, where NCA and OCA are not reliable in metallic systems.
  • The steady-state QTCI machinery could plausibly be paired with an electron-boson extension such as the Lang-Firsov transformation to treat coupled electron-phonon problems in the same framework.
  • Since the paper notes that a suitable choice of initial pivots is required in the current QTCI implementation, transferring the method to a new model should include a direct check of the fitted integrand against a sparse-grid evaluation; a more robust global pivot search would be needed for higher-dimensional or higher-order applications.
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

3 major / 5 minor

Summary. The manuscript develops a quantics tensor cross interpolation (QTCI) based impurity solver for the one-crossing approximation (OCA) in the pseudoparticle hybridization expansion, targeting equilibrium and nonequilibrium steady-state problems. The OCA self-energy integrands on a ring-shaped real-time contour are represented as quantics tensor trains and integrated directly in frequency space, and the physical Green's functions are obtained from a QTCI-fitted triangular vertex integrand. The solver is benchmarked against exact single-bath integration, a Matsubara-axis code, conventional NCA/OCA solutions on the L-shaped Kadanoff-Baym contour, and inchworm Monte Carlo data, and is applied to equilibrium and photodoped nonequilibrium DMFT spectral functions.

Significance. If the accuracy and efficiency claims hold, this is a valuable step toward practical higher-order strong-coupling solvers for nonequilibrium DMFT. The paper's strengths include the use of independent external benchmarks (exact integration, Matsubara code, L-shaped contour, inchworm data), the use of the open-source xfac QTCI library, and a transparent description of the ring-contour parametrization. The reported four-digit fitting accuracy for both smooth and discontinuous integrands is convincing in the tested fast-decay regimes. However, the central claim is broader: the metallic low-damping DMFT results rest on an untested t_max-truncation assumption, so the accuracy claim is not yet fully supported in the regime where the solver is argued to matter most.

major comments (3)
  1. [Sec. II C and Sec. III D 1] The t_max-truncation assumption in Eq. (24) is not tested. The text states that 'the pp Green's functions and hybridization functions are sufficiently decayed for relative times ≳ tmax,' and in Sec. III D 1 it asserts that t_max ≈ 6000 is 'long enough that all the pp functions are fully decayed at relative time tmax, even for the smallest considered damping.' No t_max-convergence study is provided. In the metallic benchmark, the reference OCA calculation itself was restricted to t_max = 64, and the residual discrepancy in A(ω=0) is attributed to that limited window. For η = 0.0005, a singly-occupied pp Green's function decaying like e^{-ηt} is e^{-3} ≈ 0.05 of its initial value at t = 6000, so 'fully decayed' is not obvious. This leaves the accuracy of the OCA solver in the low-η metallic regime unsupported. Please add a t_max-convergence test (e.g., A(0) for t_max = 3000, 6000, 12000 at η = 0.0005) and, if possible, a reference calculation with a larger time window.
  2. [Sec. III A and Sec. III C] The claimed 'four digits accuracy' is demonstrated for the QTCI interpolation error (σ_reference − σ_QTCI) at fixed t_max, not for the total error in the self-energy or physical Green's function. The total error also includes the truncation of the time window, which is not quantified for the metallic case. The single-bath test (η = 0.1) and the AIM benchmark (η = 0.01) use fast-decaying functions where truncation is benign, but the abstract's general accuracy claim is not supported in the small-η metallic regime. Please either report the total error against a converged reference or explicitly narrow the accuracy claim.
  3. [Sec. IV] The 'roughly two orders of magnitude' speedup relative to conventional quadrature is stated without a direct benchmark. The paper reports that one DMFT iteration takes about 10 minutes on a single processor, but no timing for the conventional OCA implementation on the same problem is given. Since efficiency is a central advertised advantage, please provide a side-by-side timing or a more detailed comparison supporting the two-order-of-magnitude claim.
minor comments (5)
  1. [Fig. 12 caption] There is a typo, 'Comparision', which should be 'Comparison'.
  2. [Sec. II B, Fig. 4] The description of the fused-leg representation is terse; a few sentences clarifying how bits of the same scale are fused and why this improves the fit of discontinuous lesser-component integrands would improve accessibility for readers unfamiliar with quantics tensor trains.
  3. [Sec. II B, Eq. (22)] The cost expression O(N_ω R (2n−1) N_bond^3) includes a factor (2n−1) whose origin is not explained; please define or derive this factor briefly.
  4. [Fig. 10] The η values are encoded only by color intensity ('from faint to vivid'), which is hard to distinguish in print; please add line styles or labels to the curves.
  5. [Sec. IV] A code availability or data availability statement would be helpful, given the use of the open-source xfac library and the reproducibility-oriented presentation.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the QTCI solver's accuracy is established against independent benchmarks, not by construction.

full rationale

The paper's derivation is self-contained: the pp self-energy is obtained by QTCI factorization of the integrand sigma(v1,v2,v3) and subsequent explicit integration via Eq. (22), and the physical Green's function is computed from Eq. (24) using the factorized gvert. No output quantity is used to define its own input: the QTCI fit is checked against exact numerical integration of the same integrand (Figs. 5 and 6), and the final self-energies and spectral functions are benchmarked against an exact single-bath solution, a Matsubara-axis code, conventional NCA/OCA on the L-shaped contour, and inchworm data from Ref. [33]. The self-citations (Refs. [8,15,22]) supply the pseudoparticle formalism, the conventional reference implementation, and the quantics encoding; they do not function as an unverified premise that forces the reported result. The tmax truncation assumption in Sec. II C is an uncontrolled numerical approximation that could affect accuracy at small eta, but it is not a circular definition and does not reduce the claimed output to the input.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

The method introduces no new physical entities; it rests on established pseudoparticle and QTCI frameworks and on numerical inputs chosen by hand. The damping parameter eta, the contour window tmax, the quantics bit count R, and the initial pivot placement are human-chosen and affect accuracy, so they are listed as free parameters. The axioms are the standard physical and numerical assumptions of the pp formalism and the OCA truncation.

free parameters (4)
  • Damping parameter eta = 0.1, 0.01, 0.005, 0.001, 0.0005, 0.0001
    Introduced in Eq. (8) to regularize the pp Green's function divergence; it broadens spectra and must be chosen small for accurate quasi-particle peaks, while small eta requires large tmax.
  • Initial QTCI pivots = six hand-chosen pivots near v1=tmax and v1+v2=tmax
    The QTCI fit quality and thus the accuracy of the final self-energy depend on the manual placement of initial pivots near the integrand discontinuities; the paper states this requirement explicitly in Secs. III A and IV.
  • Contour time window tmax = 84.7 (single bath), about 6000 (DMFT)
    The contour time window must be larger than the decay time of pp Green's functions and hybridization functions; the paper uses different tmax values per simulation, and the choice affects the truncation error.
  • Number of quantics bits R = 6 (2^18 time points)
    The number of binary digits in the quantics encoding sets the time resolution; R=6 is used in the tests, and increasing R raises the cost.
assumptions (5)
  • domain assumption Pseudoparticle mapping and the pp Dyson equation
    The solver is built on the pseudoparticle formalism of Refs. [14,15,25,26], including projection of unphysical pp states by restricting pp Green's functions to forward propagation along the contour (Sec. II A).
  • domain assumption Steady-state decoupling of the Matsubara branch
    The paper assumes memory of the initial equilibrium state is washed out so that the Matsubara branch can be pushed to t -> -infinity and only relative times matter (Sec. II A). This fails for transient dynamics.
  • domain assumption Decay of pp Green's functions and hybridization functions within tmax
    The ring-shaped contour is truncated at tmax, and the paper assumes the correlation functions have decayed for relative times exceeding tmax (Sec. II C). This is the load-bearing assumption for the time-window truncation.
  • domain assumption OCA truncation at order n=2
    The solver computes the one-crossing approximation, which is an uncontrolled truncation of the hybridization expansion; the comparison with inchworm data in Fig. 12 shows sizeable discrepancies in quasi-particle weight.
  • domain assumption QTCI low-rank compressibility of the integrand
    The method assumes the integrand sigma(v1,v2,v3) and gvert(v1,v2,v3) admit an accurate low-rank tensor-train representation; this is validated empirically for the tested parameter sets but not proven generally (Sec. III).

how reviews work

0 comments
Cite this review

Pith. "Pith review of Strong coupling impurity solver based on quantics tensor cross interpolation." pith.science (2026). https://pith.science/paper/WYTVQM5X

@misc{pith2026241119026,
  author       = {Pith},
  title        = {Pith review of: Strong coupling impurity solver based on quantics tensor cross interpolation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WYTVQM5X}},
  note         = {Machine review of arXiv:2411.19026}
}
read the original abstract

Numerical methods capable of handling nonequilibrium impurity models are essential for the study of transport problems and the solution of the nonequilibrium dynamical mean field theory (DMFT) equations. In the strong correlation regime, the self-consistently resummed hybridization expansion is an appealing strategy, which however has been employed so far mainly in the lowest-order noncrossing approximation. At higher orders, standard implementations become numerically costly, but a significant speed-up can be achieved by evaluating multidimensional integrals in an approximate factorized form. Here we develop a one-crossing approximation solver based on the recently introduced quantics tensor cross interpolation, and demonstrate its accuracy and efficiency with applications to the Anderson impurity model and nonequilibrium steady-state DMFT calculations for the Hubbard model.

Figures

Figures reproduced from arXiv: 2411.19026 by the authors.

Figure 1
Figure 1. Schematic examples of pp greater and lesser [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 3
Figure 3. OCA diagrams for the pp self-energy on the ring [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figure 4
Figure 4. Tensor-train representation of σ(v1, v2, v3). The top panel illustrates the function in the quantics represen￾tation. The middle panel shows the tensor train obtained by the QTCI factorization, and the bottom panel the tensor train with fused legs. tensor form by the QTCI method [23] ( [PITH_FULL_IMAGE:figures/full_fig_p004_4.png] view at source ↗
Figures from the paper (6 more)
Figure 5
Figure 5. Figure 5: Retarded and lesser components of the one-shot [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]
Figure 6
Figure 6. Figure 6: Greater and lesser components of the integrand of the one-shot OCA pp self-energy for the single bath model with [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 7
Figure 7. Figure 7: Pseudoparticle Green’s functions of the Ander [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]
Figure 8
Figure 8. Figure 8: Local density matrix of the Anderson impurity [PITH_FULL_IMAGE:figures/full_fig_p008_8.png]
Figure 10
Figure 10. Figure 10: Equilibrium spectral function of the half-filled [PITH_FULL_IMAGE:figures/full_fig_p009_10.png]
Figure 12
Figure 12. Figure 12: Comparision between NCA and OCA, and benchmark against the inchworm results from Ref. [33] for the doublon density nph = 0.06. Here, the parameters are T = Teff = 1/25√ 2, U = 4/ √ 2, and η = 0.0001. To enable a direct comparision with Ref. [33], the nonequilibrium di…

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Role of phonon coupling in driving photo-excited Mott insulators towards a transient superconducting steady state

    cond-mat.str-el 2024-12 conditional novelty 6.0 of 10

    Phonon cooling stabilizes a long-lived prethermal eta-paired superconducting-like state in photodoped Mott insulators, and steady-state DMFT reproduces its spectral properties.

Reference graph

Works this paper leans on

41 extracted references · 22 canonical work pages · cited by 1 Pith paper

  1. [1]

    Equilibrium system Figure 10 plots the NCA and OCA equilibrium spec- tral functions A(ω) = − 1 π ImGR(ω) for U = 2, T = 0.1 and different damping parameters η. For small η, the so- lutions from the real-time/frequency solver approach the reference data obtained with a conventional implemen- tation of the NCA and OCA solutions on the L-shaped Kadanoff-Baym...

    work page

  2. [2]

    photodoped

    Nonequilibrium system Finally, we show in Fig. 11 the nonequilibrium steady- state solutions of “photodoped” Mott insulators with an 4 2 0 2 40.00 0.25 0.50A( ) = ± 1.7 AOCA( ) A < OCA( ) 4 2 0 2 40.00 0.25 0.50A( ) = ± 2.0 fneq( )AOCA( ) 4 2 0 2 4 0.00 0.25 0.50A( ) = ± 2.3 Figure 11. OCA spectral function of the photodoped Mott insulator with U = 4 and ...

    work page

  3. [3]

    Weichselbaum and J

    A. Weichselbaum and J. von Delft, Sum-rule conserv- ing spectral functions from the numerical renormalization group, Phys. Rev. Lett. 99, 076402 (2007)

    2007

  4. [4]

    P. W. Anderson, Localized magnetic states in metals, Phys. Rev. 124, 41 (1961)

    1961

  5. [5]

    Georges, G

    A. Georges, G. Kotliar, W. Krauth, and M. J. Rozen- berg, Dynamical mean-field theory of strongly correlated fermion systems and the limit of infinite dimensions, Rev. Mod. Phys. 68, 13 (1996)

    1996

  6. [6]

    Bauernfeind, M

    D. Bauernfeind, M. Zingl, R. Triebl, M. Aichhorn, and H. G. Evertz, Fork tensor-product states: Efficient mul- tiorbital real-time dmft solver, Phys. Rev. X 7, 031013 (2017)

    work page 2017

  7. [7]

    Bulla, T

    R. Bulla, T. A. Costi, and T. Pruschke, Numerical renor- malization group method for quantum impurity systems, Rev. Mod. Phys. 80, 395 (2008)

    2008

  8. [8]

    E. Gull, A. J. Millis, A. I. Lichtenstein, A. N. Rubtsov, M. Troyer, and P. Werner, Continuous-time monte carlo methods for quantum impurity models, Rev. Mod. Phys. 83, 349 (2011)

    2011

Show all 41 references
  1. [9]

    Werner, T

    P. Werner, T. Oka, and A. J. Millis, Diagrammatic monte carlo simulation of nonequilibrium systems, Phys. Rev. B 79, 035320 (2009)

    2009

  2. [10]

    Y. Lu, X. Cao, P. Hansmann, and M. W. Haverkort, Natural-orbital impurity solver and projection approach for green’s functions, Phys. Rev. B 100, 115134 (2019)

    2019

  3. [11]

    A. J. Kim, J. Li, M. Eckstein, and P. Werner, Pseudopar- ticle vertex solver for quantum impurity models, Phys. Rev. B 106, 085124 (2022)

    2022

  4. [12]

    Cohen, E

    G. Cohen, E. Gull, D. R. Reichman, and A. J. Millis, Green’s functions from real-time bold-line monte carlo calculations: Spectral properties of the nonequilibrium anderson impurity model, Phys. Rev. Lett. 112, 146802 (2014)

    2014

  5. [13]

    Gobert, C

    D. Gobert, C. Kollath, U. Schollw¨ ock, and G. Sch¨ utz, Real-time dynamics in spin- 1 2 chains with adaptive time- dependent density matrix renormalization group, Phys. Rev. E 71, 036102 (2005)

    2005

  6. [14]

    F. B. Anders and A. Schiller, Real-time dynamics in quantum-impurity systems: A time-dependent numeri- cal renormalization-group approach, Phys. Rev. Lett.95, 196801 (2005)

    2005

  7. [15]

    Eckstein and P

    M. Eckstein and P. Werner, Nonequilibrium dynamical mean-field calculations based on the noncrossing approx- imation and its generalizations, Phys. Rev. B 82, 115115 (2010)

    2010

  8. [16]

    N´ u˜ nez Fern´ andez, M

    Y. N´ u˜ nez Fern´ andez, M. Jeannin, P. T. Dumitrescu, T. Kloss, J. Kaye, O. Parcollet, and X. Waintal, Learning feynman diagrams with tensor trains, Phys. Rev. X 12, 041018 (2022)

    2022

  9. [17]

    The nonequilibrium distribution function is chosen as fneq.(ω) = fTeff (ω − µ±) for ω ≷ 0

    one finds ∆ <(ω) = −2i( D 2 )2fneq(ω)ImGR(ω), and similarly for ∆ >(ω). The nonequilibrium distribution function is chosen as fneq.(ω) = fTeff (ω − µ±) for ω ≷ 0. The precise switching behavior near ω = 0 is not important, since we consider gapped systems. Here, fTeff (ω − µ±)...

  10. [18]

    Keiter and J

    H. Keiter and J. C. Kimball, Diagrammatic Approach to the Anderson Model for Dilute Alloys, Journal of Applied Physics 42, 1460 (1971)

    1971

  11. [19]

    Pruschke and N

    T. Pruschke and N. Grewe, The anderson model with finite coulomb repulsion, Z. Physik B - Condensed Matter 74, 439 (1989)

    1989

  12. [20]

    H. Aoki, N. Tsuji, M. Eckstein, M. Kollar, T. Oka, and P. Werner, Nonequilibrium dynamical mean-field theory and its applications, Rev. Mod. Phys. 86, 779 (2014)

    2014

  13. [21]

    Oseledets and E

    I. Oseledets and E. Tyrtyshnikov, Tt-cross approxima- tion for multidimensional arrays, Linear Algebra and its Applications 432, 70 (2010)

    2010

  14. [22]

    Savostyanov and I

    D. Savostyanov and I. Oseledets, Fast adaptive interpola- tion of multi-dimensional arrays in tensor train format, in The 2011 International Workshop on Multidimensional (nD) Systems (2011) pp. 1–8

    2011

  15. [23]

    D. V. Savostyanov, Quasioptimality of maximum-volume cross interpolation of tensors, Linear Algebra and its Ap- plications 458, 217 (2014)

    2014

  16. [24]

    Matsuura, H

    S. Matsuura, H. Shinaoka, P. Werner, and N. Tsuji, Tensor cross interpolation approach for quantum im- purity problems based on the weak-coupling expansion, arXiv:2501.12643 (2025)

    2025 arXiv

  17. [25]

    Shinaoka, M

    H. Shinaoka, M. Wallerberger, Y. Murakami, K. Nogaki, R. Sakurai, P. Werner, and A. Kauch, Multiscale space- time ansatz for correlation functions of quantum systems based on quantics tensor trains, Phys. Rev. X 13, 021015 (2023)

    2023

  18. [26]

    M. K. Ritter, Y. N´ u˜ nez Fern´ andez, M. Wallerberger, J. von Delft, H. Shinaoka, and X. Waintal, Quantics tensor cross interpolation for high-resolution parsimo- nious representations of multivariate functions, Phys. Rev. Lett. 132, 056501 (2024)

    2024

  19. [27]

    Li and M

    J. Li and M. Eckstein, Nonequilibrium steady-state the- ory of photodoped mott insulators, Phys. Rev. B 103, 045133 (2021)

    2021

  20. [28]

    Coleman, New approach to the mixed-valence prob- lem, Phys

    P. Coleman, New approach to the mixed-valence prob- lem, Phys. Rev. B 29, 3035 (1984). 12

    1984

  21. [29]

    N. E. Bickers, Review of techniques in the large-n expan- sion for dilute magnetic alloys, Rev. Mod. Phys. 59, 845 (1987)

    1987

  22. [30]

    Yuriel Nunez-Fernandez, M

    Y. Yuriel Nunez-Fernandez, M. Ritter, M. Jeannin, J.-W. Li, T. Kloss, T. Louvet, S. Terasaki, O. Parcollet, J. von Delft, H. Shinaoka, and X. Waintal, Learning tensor net- works with tensor cross interpolation: new algorithms and libraries, arXiv:2407.02454v2 (2024)

    2024 arXiv

  23. [31]

    We use the code implemented for benchmark purposes in Ref

    Note that the OCA calculation on the Matsubara axis does not involve the vertex self-consistency. We use the code implemented for benchmark purposes in Ref. [8]

  24. [32]

    Bl¨ umer,Mott-Hubbard Metal-Insulator Transition and Optical Conductivity in High Dimensions, Ph.D

    N. Bl¨ umer,Mott-Hubbard Metal-Insulator Transition and Optical Conductivity in High Dimensions, Ph.D. thesis (2000)

    2000

  25. [33]

    J. Li, D. Golez, P. Werner, and M. Eckstein, η-paired superconducting hidden phase in photodoped mott insu- lators, Phys. Rev. B 102, 165136 (2020)

    2020

  26. [34]

    S. Ray, Y. Murakami, and P. Werner, Nonthermal su- perconductivity in photodoped multiorbital hubbard sys- tems, Phys. Rev. B 108, 174515 (2023)

    2023

  27. [35]

    Ray and P

    S. Ray and P. Werner, Photoinduced ferromagnetic and superconducting orders in multiorbital hubbard models, Phys. Rev. B 110, L041109 (2024)

    2024

  28. [36]

    K¨ unzel, A

    F. K¨ unzel, A. Erpenbeck, D. Werner, E. Arrigoni, E. Gull, G. Cohen, and M. Eckstein, Numerically ex- act simulation of photodoped mott insulators, Phys. Rev. Lett. 132, 176501 (2024)

    2024

  29. [37]

    Thoenniss, A

    J. Thoenniss, A. Lerose, and D. A. Abanin, Nonequi- librium quantum impurity problems via matrix-product states in the temporal domain, Phys. Rev. B 107, 195101 (2023)

    2023

  30. [38]

    Goleˇ z, M

    D. Goleˇ z, M. Eckstein, and P. Werner, Dynamics of screening in photodoped mott insulators, Phys. Rev. B 92, 195123 (2015)

    2015

  31. [39]

    Werner and M

    P. Werner and M. Eckstein, Phonon-enhanced relaxation and excitation in the holstein-hubbard model, Phys. Rev. B 88, 165108 (2013)

    2013

  32. [40]

    H. U. R. Strand, D. Goleˇ z, M. Eckstein, and P. Werner, Hund’s coupling driven photocarrier relaxation in the two-band mott insulator, Phys. Rev. B 96, 165104 (2017)

    2017

  33. [41]

    Eckstein, Solving quantum impurity models in the non-equilibrium steady state with tensor trains, arXiv:2410.19707 (2024)

    M. Eckstein, Solving quantum impurity models in the non-equilibrium steady state with tensor trains, arXiv:2410.19707 (2024)

    2024 arXiv

Pith tools

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