Pith. sign in

REVIEW 3 major objections 4 minor 74 references

This paper establishes that transcorrelated integrals have a low-rank structure that interpolative separable density-fitting can exploit, making transcorrelated coupled-cluster calculations practical for systems with over a thousand orbital

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · deepseek-v4-flash

2026-08-01 18:21 UTC pith:SQRPMVG3

load-bearing objection Solid methods paper that makes transcorrelated CCSD practical, but the production benchmarks need a direct ISDF rank-convergence check and the SI is missing. the 3 major comments →

arxiv 2607.17314 v1 pith:SQRPMVG3 submitted 2026-07-19 physics.chem-ph cond-mat.str-elphysics.comp-ph

Interpolative Separable Density-Fitting for Transcorrelated Hamiltonians

Ke Liao , Yifan Cheng , Werner Dobrautz , Tianyu Zhu , Ali Alavi This is my paper
classification physics.chem-ph cond-mat.str-elphysics.comp-ph
keywords transcorrelationinterpolative separable density fittingcoupled clustercomplete-basis-set limitJastrow factorlow-rank approximationGPU accelerationhydrogen chain
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

The paper aims to make transcorrelated methods—which fold a Jastrow correlation factor into the Hamiltonian and thereby accelerate convergence to the complete-basis-set limit—feasible for large molecules. Its central insight is that the transcorrelated two-body and effective two-body integrals, previously a computational bottleneck when evaluated on a numerical grid, inherit a low-rank structure from the orbital-pair densities. The authors compress these integrals with interpolative separable density-fitting (ISDF) and combine this with the xTC reduction of the three-body operator, cutting integration and storage costs by orders of magnitude. They demonstrate the approach on hydrogen chains, matching state-of-the-art many-body references to within about 1 mHa/atom in the thermodynamic limit, and on benzene with up to 1200 orbitals (cc-pCV5Z), where it captures a correlation energy close to neural-network variational Monte Carlo at the formal cost of CCSD.

Core claim

The core discovery is that the transcorrelated K integrals and the effective two-body ΔU correction of the xTC scheme can be factorized through ISDF: products of orbitals and orbital gradients on a real-space grid are approximated by a small set of interpolation points with auxiliary basis functions. Because the rank of the contracted four-index tensor is bounded by the numerical rank of the orbital-pair density, a per-channel rank of N_ch = 12N_orb (fused to N_mu ≈ 1.8–2.2 N_ch) suffices in production. The pivoted-Cholesky selection of interpolation points depends only on the orbital products, not on the Jastrow factor, and automatic differentiation of the correlator eliminates hand-coded i

What carries the argument

The load-bearing ingredient is the ISDF factorization of orbital-pair products on a numerical grid: rho_p^q(r) ≈ Σ_mu xi_mu(r) C_p^q(r_mu), with a parallel factorization for orbital-gradient products. The interpolation points r_mu are chosen by a matrix-free pivoted-Cholesky decomposition of the overlap Gram matrix, independent of the Jastrow; scalar and gradient pivot sets are fused so that all contractions share a single auxiliary index. The effective two-body ΔU is built from four terms, two of which retain an open orbital index and are made tractable by a low-rank projector derived from the idempotent reference density matrix. This machinery replaces the dense N_g^2 grid-pair sums with O

Load-bearing premise

The fixed per-channel ISDF rank of 12 times the number of orbitals, validated only on ethylene, is assumed to keep the truncation error below the claimed accuracy for the larger, more diffuse hydrogen-chain and benzene systems, without a direct convergence check on those production systems.

What would settle it

For a molecule such as benzene in a diffuse basis or H50 at cc-pV5Z, compute the ISDF-xTC-CCSD total energy at N_ch = 12N_orb and compare it with the same calculation at N_ch = 20N_orb or with a dense non-ISDF reference; if the per-atom energy difference exceeds the reported accuracy (e.g., more than 1 mHa/atom or the claimed sub-mHa deviations), the production rank is inadequate.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • Transcorrelated CCSD at a triple-zeta basis already surpasses canonical CCSD at a quintuple-zeta basis for the hydrogen chain, effectively saving about two cardinal numbers.
  • The ISDF truncation error is systematically convergent: for ethylene in a triple-zeta basis it falls below 1 mHa by N_ch ≈ 6N_orb and below 0.01 mHa by 10N_orb, with the production choice of 12N_orb comfortably converged.
  • The method remains tractable at 1200 orbitals (benzene cc-pCV5Z), with the ISDF setup time scaling empirically as N_orb^1.76 over the studied range.
  • The CBS extrapolation of the transcorrelated correlation energy is markedly less sensitive to the choice of exponential versus inverse-power fit form (1.1 mHa spread) than conventional CCSD and CCSD(T) (7.6 and 8.0 mHa).
  • Because the ISDF pivots are selected from the orbital products alone, independent of the Jastrow, the same construction works for any differentiable correlator without changing the pivot-selection procedure.

Where Pith is reading between the lines

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

  • The fixed per-channel rank N_ch = 12N_orb was validated only on ethylene; if the numerical rank of the orbital-pair density grows with system size or diffuseness, the production systems could carry larger ISDF errors than reported, although the agreement with high-level references suggests the errors remain small.
  • Because the pivots are Jastrow-independent, the ISDF factorization could be reused across multiple Jastrow forms or during iterative Jastrow optimization, amortizing the setup cost.
  • The factor-direct CCSD formulation flagged as under development would avoid explicit construction of the four-virtual integral tensor and lower the dominant contraction from nominal sixth- to fifth-order scaling, extending the method to yet larger systems.
  • The low-rank behavior of the transcorrelated integrals likely persists in periodic plane-wave settings, where the orbital-pair density is even more structured, making the approach a promising route to transcorrelated solids.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

3 major / 4 minor

Summary. The paper introduces an ISDF-based construction of transcorrelated Hamiltonians for general differentiable Jastrow factors. The transcorrelated two-body kernel K and the xTC effective two-body correction ΔU are factored through orbital-pair and orbital-gradient density fitting, with pivoted-Cholesky pivot selection and GPU/out-of-core implementation. The method is benchmarked on the linear hydrogen chain, where TDL xTC-CCSD energies are compared with AFQMC+ΔDMRG references, and on benzene up to cc-pCV5Z (1200 orbitals), where the CBS correlation energy is compared with CCSD, CCSD(T), and neural-network VMC/DMC references. The central claim is that ISDF compression makes transcorrelated integral construction practical while preserving sub-mHa accuracy at the production rank.

Significance. If the production-rank validation is supplied, this would be a valuable contribution: it addresses a real bottleneck in transcorrelated calculations and combines several non-trivial ingredients — ISDF for TC/xTC integrals, automatic differentiation of Jastrow factors, a matrix-free pivot search, and a multi-GPU out-of-core pipeline. The formal derivations in §III are careful and internally consistent, and the ethylene rank-convergence test in Fig. 1 is an appropriate external check against a dense non-ISDF reference. The explicit code and data availability statements are a further strength. The main unresolved issue is whether the single-system ISDF convergence test transfers to the larger, more diffuse production systems; the manuscript also lacks the supporting information it repeatedly cites, which prevents full verification of the TDL and CBS claims.

major comments (3)
  1. [§VI.A, Fig. 1; Table IV] The adequacy of the production ISDF rank N_ch = 12 N_orb is established only for ethylene in cc-pVTZ (N_orb = 116). The H-chain and benzene production runs (Table IV, N_orb up to 525 and 1200) use the same fixed per-channel rank without a direct convergence check against a dense reference or a higher-rank calculation. Because Fig. 1 shows that rank ratios as low as 4 N_orb are under-resolved by ~18 mHa, the transferability of the 12 N_orb setting to larger and more diffuse basis sets is not guaranteed. This is load-bearing: the reported sub-mHa H-chain agreement and the benzene CBS value are energies of the ISDF-truncated Hamiltonian. Please add rank-convergence data for at least one production system (e.g., H10 or H20 in the cc-pV5Z MP2NO space and benzene in cc-pCVQZ), reporting total-energy and/or K/ΔU integral errors, and state how the TDL/CBS values move with N_ch.
  2. [§II.C–§VI.C (placeholder SI references)] The manuscript repeatedly refers to 'Supporting Information Section??', 'Fig.??', and 'Table??' for material needed to verify central results: the VMC optimization schedule (walker counts, learning rates, averaging windows), the CCSD block-tiling and streaming protocol, the per-separation TDL fits, and the CBS fit parameters. As submitted, these data are absent, so claims such as the 1/N TDL extrapolation and the benzene CBS value cannot be independently checked. This is more than a formatting issue and must be fixed before acceptance.
  3. [§VI.B, Figs. 2 and 4] The production cc-pV5Z TDL series in Fig. 4 (open stars) uses the MP2NO cutoff n_keep/N_atom = 10 directly, whereas the cutoff validation in Fig. 2 is a single-geometry (r = 1.4 bohr) three-point fit for H10. The text itself notes that this fit has no independent statistical error and can retain fit-form bias. Since the residual per-atom error from the cutoff is inferred to be ~0.6 mHa/atom for the TDL series, please provide per-separation cutoff convergence data or an error bound, and clarify whether the open-star series is a full-virtual result or a fixed-truncation result. The current wording 'joint thermodynamic and CBS limits' goes beyond what is documented.
minor comments (4)
  1. [Abstract and §VI.B] The phrase 'joint thermodynamic and CBS limits' overstates the H-chain treatment: the cc-pV5Z production series uses an MP2NO truncation (n_keep/N_atom = 10) and the TDL is obtained by linear 1/N fits. Please qualify the wording to distinguish full-basis/full-virtual limits from fixed-truncation production results.
  2. [§III.A, Eq. (25)] In Eq. (25), the gradient auxiliary factor is written as \tilde{C}^p_r(r_\mu) without the Cartesian component index c that appears in Eq. (18). Adding the explicit c summation would remove ambiguity in the subsequent contraction with u^{(1),c}_{\mu\nu}.
  3. [§III.B, Eqs. (37)–(38)] The claimed cost reduction for Terms 2 and 3 depends on the rank α of the L_Q projector. For the closed-shell HF reference, α = N_occ, which is O(N_orb). Please state the α dependence explicitly in the scaling summary and provide a wall-clock breakdown for the measured five-fold speedup, so the reader can see which contraction is actually rate-limiting.
  4. [§VI.C, Fig. 6] The setup-time scaling T[s] ≈ 0.030 N_orb^1.76 is a three-point power-law fit and is described as such in the text; the caption and figure would benefit from an explicit statement that this is not an asymptotic scaling law, to avoid over-interpretation.

Circularity Check

0 steps flagged

No significant circularity: ISDF is validated against a dense non-ISDF reference, and the benchmark targets are external many-body results.

full rationale

The central derivation—ISDF compression of the transcorrelated K and ΔU integrals—does not reduce to its own inputs. The approximation is validated in Fig. 1 on ethylene against an explicit dense non-ISDF transcorrelated reference with the same Jastrow and solver, and the paper reports that the ISDF truncation error falls below 1 mHa by Nch≈6Norb and is converged to <0.01 mHa by Nch=10Norb. This is an independent, non-circular accuracy check of the compression scheme. The hydrogen-chain thermodynamic-limit energies are compared with external AFQMC+ΔDMRG data of Motta et al., and the benzene CBS energies are compared with external CCSD(T), FermiNet-VMC/DMC, and DMC benchmarks; no fitted parameter is renamed as a prediction. The xTC treatment and non-Hermitian CC solver are imported from prior overlapping work (Refs. 17–19, 23), but the paper re-derives the contraction algebra in Eqs. (25)–(40) and does not use those citations to force the reported accuracy; no uniqueness theorem is invoked. The use of the fixed rank Nch=12Norb in production systems, validated only on ethylene, and the missing SI details flagged by the manuscript itself ('Supporting Information Section??' in Sections II.C, IV.A, IV.C, V.b, V.c, and VI) are correctness/completeness limitations, not circularity. No equation in the derivation is equivalent to its inputs by construction.

Axiom & Free-Parameter Ledger

4 free parameters · 7 axioms · 0 invented entities

The paper introduces no new physical entities. The free parameters are numerical convergence settings and ansatz choices; the axioms are standard mathematical facts plus domain assumptions from prior work. The most fragile implicit assumption is that the ISDF rank and grid settings transfer from small benchmarks to the production systems.

free parameters (4)
  • ISDF rank ratio Nch/Norb = 12
    Chosen by hand; validated on ethylene (Fig. 1) but not on production systems.
  • MP2NO virtual truncation n_keep/N_atom = 10
    Production cutoff for H chains; error <0.12 mHa/atom on H10, assumed for longer chains.
  • Boys–Handy correlator exponent set = 17 terms per nucleus type; exponents (mk,nk,ok)
    Ansatz choice; increases expressivity but adds parameters.
  • VMC clipping factor and Gauss–Newton Tikhonov shift = 5σ; λ adaptive
    Heuristic stability parameters; default choices not derived.
axioms (7)
  • standard math The Baker–Campbell–Hausdorff expansion of the similarity transformation terminates at the second commutator, yielding at most three-body operators (Eq. 3).
    Position-only τ with no derivative operators makes higher commutators vanish; this is standard for transcorrelated Hamiltonians.
  • domain assumption The surface terms in the integration by parts of K(2) vanish for square-integrable localized orbital products (Eq. 23).
    Assumes orbital products decay fast enough at the grid boundary; standard for molecular calculations.
  • standard math The rank-bound rank(AB) ≤ min(rank(A), rank(B)) ensures K, ΔU, and L have numerical rank ≤ Nμ (Sec II D).
    Justifies the ISDF low-rank representation; relies on the orbital-overlap density being low rank.
  • domain assumption The xTC approximation (contracting the three-body operator with the reference density matrix) is accurate for the systems studied (Ref. 23).
    Prior work establishes this; the present paper does not benchmark xTC vs full TC.
  • domain assumption The PySCF level-2 Becke grid is sufficiently accurate for TC integrals (Sec V a, adopted from Ref. 21).
    Grid convergence not re-tested here; N_g is fixed at 99,480 for all benzene bases.
  • domain assumption MP2 natural-orbital truncation at 10 virtuals per atom introduces errors below 0.12 mHa/atom for the hydrogen chain (Sec VI B).
    Validated on H10; assumed to transfer to longer chains.
  • domain assumption Thermodynamic-limit and CBS extrapolations are described by linear 1/N and exponential/inverse-power forms respectively (Sec VI B,C).
    Standard empirical fits; sensitivity checks reported in SI, not main text.

pith-pipeline@v1.3.0-alltime-deepseek · 24034 in / 15934 out tokens · 156812 ms · 2026-08-01T18:21:29.675109+00:00 · methodology

0 comments
read the original abstract

The transcorrelated (TC) method dramatically accelerates the convergence of correlated calculations toward the complete-basis-set (CBS) limit by folding a Jastrow correlator into the Hamiltonian via a similarity transformation, incorporating the electron--electron cusp into the effective interaction. We make the TC framework practical for large systems and flexible, multi-center correlators by compressing the grid-evaluated TC integrals with the interpolative separable density-fitting (ISDF) approximation, combined with the effective two-body (xTC) treatment of the three-body operator. This low-rank representation reduces storage and integration costs by orders of magnitude, and a multi-GPU implementation with automatic differentiation of the correlator makes the construction routine for large basis sets. We demonstrate the resulting ISDF-xTC-CCSD method on the linear hydrogen chain, reaching the joint thermodynamic and CBS limits with basis sets up to cc-pV5Z in agreement with state-of-the-art many-body references to within about 1~mHa/atom, and on the benzene ground-state energy with up to 1200 orbitals (cc-pCV5Z), where the method attains state-of-the-art accuracy at the coupled cluster singles and doubles level and its CBS extrapolation is markedly more robust than that of conventional coupled-cluster methods.

Figures

Figures reproduced from arXiv: 2607.17314 by Ali Alavi, Ke Liao, Tianyu Zhu, Werner Dobrautz, Yifan Cheng.

Figure 1
Figure 1. Figure 1: FIG. 1: Convergence of the ISDF approximation with [PITH_FULL_IMAGE:figures/full_fig_p011_1.png] view at source ↗
Figure 3
Figure 3. Figure 3: FIG. 3: Basis-set convergence on the H [PITH_FULL_IMAGE:figures/full_fig_p012_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: FIG. 4: Deviation of the per-atom energy from the [PITH_FULL_IMAGE:figures/full_fig_p013_4.png] view at source ↗
Figure 6
Figure 6. Figure 6: FIG. 6: Basis-set scaling of the ISDF-xTC setup [PITH_FULL_IMAGE:figures/full_fig_p014_6.png] view at source ↗

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

74 extracted references · 8 canonical work pages · 1 internal anchor

  1. [1]

    Compact numerical solutions to the two-dimensional repulsive Hubbard model obtained via nonunitary similarity transformations , volume =

    Dobrautz, Werner and Luo, Hongjun and Alavi, Ali , year =. Compact numerical solutions to the two-dimensional repulsive Hubbard model obtained via nonunitary similarity transformations , volume =. Physical Review B , publisher =. doi:10.1103/physrevb.99.075119 , number =

    work page doi:10.1103/physrevb.99.075119

  2. [2]

    and Alavi, Ali and Giner, Emmanuel , year =

    Dobrautz, Werner and Cohen, Aron J. and Alavi, Ali and Giner, Emmanuel , year =. Performance of a one-parameter correlation factor for transcorrelation: Study on a series of second row atomic and molecular systems , volume =. The Journal of Chemical Physics , publisher =. doi:10.1063/5.0088981 , number =

    work page doi:10.1063/5.0088981

  3. [3]

    Real-Space Chemistry on Quantum Computers: A Fault-Tolerant Algorithm with Adaptive Grids and Transcorrelated Extension

    Feniou, César and Cherfan, Christopher and Zylberman, Julien and Claudon, Baptiste and Piquemal, Jean-Philip and Giner, Emmanuel , keywords =. Real-Space Chemistry on Quantum Computers: A Fault-Tolerant Algorithm with Adaptive Grids and Transcorrelated Extension , publisher =. 2025 , copyright =. doi:10.48550/ARXIV.2507.20583 , url =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.2507.20583 2025

  4. [4]

    , year =

    Uvarov, Alexey and Izmaylov, Artur F. , year =. Accuracy and Resource Advantages of Quantum Eigenvalue Estimation with Non-Hermitian Transcorrelated Electronic Hamiltonians , volume =. Journal of Chemical Theory and Computation , publisher =. doi:10.1021/acs.jctc.6c00274 , number =

    doi:10.1021/acs.jctc.6c00274

  5. [5]

    Towards efficient quantum computing for quantum chemistry: reducing circuit complexity with transcorrelated and adaptive ansatz techniques , volume =

    Magnusson, Erika and Fitzpatrick, Aaron and Knecht, Stefan and Rahm, Martin and Dobrautz, Werner , year =. Towards efficient quantum computing for quantum chemistry: reducing circuit complexity with transcorrelated and adaptive ansatz techniques , volume =. doi:10.1039/d4fd00039k , journal =

    work page doi:10.1039/d4fd00039k

  6. [6]

    and Dobrautz, Werner and Luo, Hongjun and Alavi, Ali and Tavernelli, Ivano , year =

    Sokolov, Igor O. and Dobrautz, Werner and Luo, Hongjun and Alavi, Ali and Tavernelli, Ivano , year =. Orders of magnitude increased accuracy for quantum many-body problems on quantum computers via an exact transcorrelated method , volume =. Physical Review Research , publisher =. doi:10.1103/physrevresearch.5.023174 , number =

    work page doi:10.1103/physrevresearch.5.023174

  7. [7]

    2023 , month = jul, journal =

    Christlmaier, Evelin Martine and Schraivogel, Thomas and L. 2023 , month = jul, journal =. doi:10.1063/5.0154445 , url =

    doi:10.1063/5.0154445 2023

  8. [8]

    and Luo, Hongjun and Guther, Kai and Dobrautz, Werner and Tew, David P

    Cohen, Aron J. and Luo, Hongjun and Guther, Kai and Dobrautz, Werner and Tew, David P. and Alavi, Ali , year =. Similarity Transformation of the Electronic. J. Chem. Phys. , volume =. doi:10.1063/1.5116024 , url =

    work page doi:10.1063/1.5116024

  9. [9]

    2021 , month = jul, journal =

    Towards Efficient and Accurate Ab Initio Solutions to Periodic Systems via Transcorrelation and Coupled Cluster Theory , author =. 2021 , month = jul, journal =. doi:10.1103/PhysRevResearch.3.033072 , url =

    doi:10.1103/physrevresearch.3.033072 2021

  10. [10]

    2023 , month = mar, journal =

    Density Matrix Renormalization Group for Transcorrelated Hamiltonians: Ground and Excited States in Molecules , author =. 2023 , month = mar, journal =. doi:10.1021/acs.jctc.2c01207 , url =

    doi:10.1021/acs.jctc.2c01207 2023

  11. [11]

    Transcorrelated Coupled Cluster Methods

    Schraivogel, Thomas and Christlmaier, Evelin Martine Corvid and. Transcorrelated Coupled Cluster Methods. 2023 , journal =. doi:10.1063/5.0151412 , url =

    doi:10.1063/5.0151412 2023

  12. [12]

    Combining the

    Luo, Hongjun and Alavi, Ali , year =. Combining the. J. Chem. Theory Comput. , volume =. doi:10.1021/acs.jctc.7b01257 , url =

    doi:10.1021/acs.jctc.7b01257

  13. [13]

    2017 , month = sep, journal =

    Towards the Solution of the Many-Electron Problem in Real Materials: Equation of State of the Hydrogen Chain with State-of-the-Art Many-Body Methods , author =. 2017 , month = sep, journal =. doi:10.1103/PhysRevX.7.031059 , url =

    doi:10.1103/physrevx.7.031059 2017

  14. [14]

    2023 , journal =

    Towards the Ground State of Molecules via Diffusion Monte Carlo on Neural Networks , author =. 2023 , journal =. doi:10.1038/s41467-023-37609-3 , url =

    work page doi:10.1038/s41467-023-37609-3 2023

  15. [15]

    2023 , booktitle =

    A Self-Attention Ansatz for Ab-initio Quantum Chemistry , author =. 2023 , booktitle =. 2211.13672 , archiveprefix =

    Pith/arXiv arXiv 2023

  16. [16]

    2023 , journal =

    Optimizing Jastrow Factors for the Transcorrelated Method , author =. 2023 , journal =. doi:10.1063/5.0147877 , url =

    doi:10.1063/5.0147877 2023

  17. [17]

    2025 , journal =

    Transcorrelated Methods for Multireference Problems , author =. 2025 , journal =. doi:10.1063/5.0282673 , url =

    work page doi:10.1063/5.0282673 2025

  18. [18]

    Boys, S. F. and Handy, N. C. , year =. A First Solution, for. Proc. R. Soc. Lond. A , volume =

  19. [19]

    1969 , journal =

    Some Bilinear Convergence Characteristics of the Solutions of Dissymmetric Secular Equations , author =. 1969 , journal =

    1969

  20. [20]

    1969 , journal =

    A Condition to Remove the Indeterminacy in Interelectronic Correlation Functions , author =. 1969 , journal =

    1969

  21. [21]

    1969 , journal =

    The Determination of Energies and Wavefunctions with Full Electronic Correlation , author =. 1969 , journal =

    1969

  22. [22]

    1969 , journal =

    A Calculation for the Energies and Wavefunctions for States of Neon with Full Electronic Correlation Accuracy , author =. 1969 , journal =

    1969

  23. [23]

    2015 , journal =

    Compression of the Electron Repulsion Integral Tensor in Tensor Hypercontraction Format with Cubic Scaling Cost , author =. 2015 , journal =

    2015

  24. [24]

    2017 , journal =

    Interpolative Separable Density Fitting Decomposition for Accelerating Hybrid Density Functional Calculations with Applications to Defects in Silicon , author =. 2017 , journal =

    2017

  25. [25]

    1988 , journal =

    A Multicenter Numerical Integration Scheme for Polyatomic Molecules , author =. 1988 , journal =

    1988

  26. [26]

    and Blunt, Nick S

    Guther, Kai and Anderson, Robert J. and Blunt, Nick S. and Bogdanov, Nikolay A. and Cleland, Deidre and Dattani, Nike and Dobrautz, Werner and Ghanem, Khaldoon and Jeszenszki, Peter and Liebermann, Niklas and Manni, Giovanni Li and Lozovoi, Alexander Y. and Luo, Hongjun and Ma, Dongxia and Merz, Florian and Overy, Catherine and Rampp, Markus and Samanta, ...

  27. [27]

    Umrigar, C. J. , year =. Accelerated. Phys. Rev. Lett. , volume =

  28. [28]

    1950 , journal =

    Adjustment of an Inverse Matrix Corresponding to a Change in One Element of a Given Matrix , author =. 1950 , journal =

    1950

  29. [29]

    and Chester, G

    Ceperley, D. and Chester, G. V. and Kalos, M. H. , year =. Monte. Phys. Rev. B , volume =

  30. [30]

    1992 , journal =

    Acceleration of Stochastic Approximation by Averaging , author =. 1992 , journal =

    1992

  31. [31]

    Kats, Daniel and Christlmaier, Evelin M. C. and Schraivogel, Thomas and Alavi, Ali , title =. Faraday Discuss. , volume =. 2024 , doi =

    2024

  32. [32]

    Transcorrelated methods applied to second row elements , journal =

    Filip, Maria-Andreea and L. Transcorrelated methods applied to second row elements , journal =. 2025 , doi =

    2025

  33. [33]

    Ammar, Abdallah and Scemama, Anthony and Giner, Emmanuel , title =. J. Chem. Theory Comput. , volume =. 2023 , doi =

    2023

  34. [34]

    Philip and Filip, Maria-Andreea and Christlmaier, Evelin M

    Haupt, J. Philip and Filip, Maria-Andreea and Christlmaier, Evelin M. C. and Cheng, Yifan and Hauskrecht, Johannes and Alavi, Ali , title =. 2025 , eprint =

    2025

  35. [35]

    Filip, Maria-Andreea and Christlmaier, Evelin M. C. and Haupt, J. Philip and Kats, Daniel and L. Deterministic Optimisation of Jastrow Factors , journal =. 2025 , eprint =

    2025

  36. [36]

    Non-iterative Triples for Transcorrelated Coupled Cluster Theory , journal =

    M. Non-iterative Triples for Transcorrelated Coupled Cluster Theory , journal =. 2025 , doi =

    2025

  37. [37]

    Simula, Kristoffer and Hauskrecht, Johannes and Christlmaier, Evelin M. C. and L. A Transcorrelated Wave-Function Framework for Solids: An Application to Bulk and Defected Silicon , journal =. 2025 , eprint =

    2025

  38. [38]

    , title =

    Ten-no, Seiichiro L. , title =. J. Chem. Phys. , volume =. 2023 , doi =

    2023

  39. [39]

    Explicitly Correlated Electronic Structure Calculations with Transcorrelated Matrix Product Operators , journal =

    Baiardi, Alberto and Lesiuk, Micha. Explicitly Correlated Electronic Structure Calculations with Transcorrelated Matrix Product Operators , journal =. 2022 , doi =

    2022

  40. [40]

    Qin, Xinming and Hu, Wei and Yang, Jinlong , title =. J. Chem. Theory Comput. , volume =. 2023 , doi =

    2023

  41. [41]

    , title =

    Kutzelnigg, Werner and Morgan, John D. , title =. J. Chem. Phys. , volume =. 1992 , doi =

    1992

  42. [42]

    Helgaker, Trygve and Klopper, Wim and Koch, Henrik and Noga, Jozef , title =. J. Chem. Phys. , volume =. 1997 , doi =

    1997

  43. [43]

    Drummond, N. D. and Towler, M. D. and Needs, R. J. , title =. Phys. Rev. B , volume =. 2004 , doi =

    2004

  44. [44]

    and Parrish, Robert M

    Hohenstein, Edward G. and Parrish, Robert M. and Mart. Tensor hypercontraction density fitting. J. Chem. Phys. , volume =. 2012 , doi =

    2012

  45. [45]

    and Zhang, Shuai and Morales, Miguel A

    Malone, Fionn D. and Zhang, Shuai and Morales, Miguel A. , title =. J. Chem. Theory Comput. , volume =. 2019 , doi =

    2019

  46. [46]

    Dong, Kun and Hu, Wei and Lin, Lin , title =. J. Chem. Theory Comput. , volume =. 2018 , doi =

    2018

  47. [47]

    and Alavi, Ali and Kats, Daniel , title =

    Schraivogel, Thomas and Cohen, Aron J. and Alavi, Ali and Kats, Daniel , title =. J. Chem. Phys. , volume =. 2021 , doi =

    2021

  48. [48]

    Giner, Emmanuel , title =. J. Chem. Phys. , volume =. 2021 , doi =

    2021

  49. [49]

    and Liao, Ke and L

    Dobrautz, Werner and Sokolov, Igor O. and Liao, Ke and L. Toward Real Chemical Accuracy on Current Quantum Hardware Through the Transcorrelated Method , journal =. 2024 , doi =

    2024

  50. [50]

    Philip and Kats, Daniel and L

    Simula, Kristoffer and Christlmaier, Evelin Martine Corvid and Filip, Maria-Andreea and Haupt, J. Philip and Kats, Daniel and L. Transcorrelated Theory with Pseudopotentials , journal =. 2025 , doi =

    2025

  51. [51]

    Kutzelnigg, Werner , title =. Theor. Chim. Acta , volume =. 1985 , doi =

    1985

  52. [52]

    Klopper, Wim and Kutzelnigg, Werner , title =. Chem. Phys. Lett. , volume =. 1987 , doi =

    1987

  53. [53]

    Ten-no, Seiichiro , title =. Chem. Phys. Lett. , volume =. 2004 , doi =

    2004

  54. [54]

    Explicitly Correlated Electrons in Molecules , journal =

    H. Explicitly Correlated Electrons in Molecules , journal =. 2012 , doi =

    2012

  55. [55]

    James Bradbury and Roy Frostig and Peter Hawkins and Matthew James Johnson and Yash Katariya and Chris Leary and Dougal Maclaurin and George Necula and Adam Paszke and Jake Vander

  56. [56]

    Berkelbach and Nick S

    Qiming Sun and Timothy C. Berkelbach and Nick S. Blunt and George H. Booth and Sheng Guo and Zhendong Li and Junzi Liu and James D. McClain and Elvira R. Sayfutyarova and Sandeep Sharma and Sebastian Wouters and Garnet Kin-Lic Chan , title =. WIREs Comput. Mol. Sci. , volume =. 2018 , doi =

    2018

  57. [57]

    Blunt and Nikolay A

    Qiming Sun and Xing Zhang and Samragni Banerjee and Peng Bao and Marc Barbry and Nick S. Blunt and Nikolay A. Bogdanov and George H. Booth and Jia Chen and Zhi-Hao Cui and Janus J. Eriksen and Yang Gao and Sheng Guo and Jan Hermann and Matthew R. Hermes and Kevin Koh and Peter Koval and Susi Lehtola and Zhendong Li and Junzi Liu and Narbe Mardirossian and...

    2020

  58. [58]

    2026 , eprint =

    An Additive Reference Correction Scheme for the Transcorrelated Method , author =. 2026 , eprint =

    2026

  59. [59]

    Umezawa, Naoto and Tsuneyuki, Shinji , title =. J. Chem. Phys. , volume =. 2003 , doi =

    2003

  60. [60]

    Sakuma, Rei and Tsuneyuki, Shinji , title =. J. Phys. Soc. Jpn. , volume =. 2006 , doi =

    2006

  61. [61]

    Ochi, Masayuki and Sodeyama, Keitaro and Sakuma, Rei and Tsuneyuki, Shinji , title =. J. Chem. Phys. , volume =. 2012 , doi =

    2012

  62. [62]

    Ochi, Masayuki and Arita, Ryotaro and Tsuneyuki, Shinji , title =. Phys. Rev. Lett. , volume =. 2017 , doi =

    2017

  63. [63]

    , title =

    Gao, Weiwei and Chelikowsky, James R. , title =. J. Chem. Theory Comput. , volume =. 2020 , doi =

    2020

  64. [64]

    Lee, Joonho and Lin, Lin and Head-Gordon, Martin , title =. J. Chem. Theory Comput. , volume =. 2020 , doi =

    2020

  65. [65]

    and Zhang, Shuai and Morales, Miguel A

    Malone, Fionn D. and Zhang, Shuai and Morales, Miguel A. , title =. J. Chem. Theory Comput. , volume =. 2020 , doi =

    2020

  66. [66]

    and Gidney, Craig and Huggins, William J

    Lee, Joonho and Berry, Dominic W. and Gidney, Craig and Huggins, William J. and McClean, Jarrod R. and Wiebe, Nathan and Babbush, Ryan , title =. PRX Quantum , volume =. 2021 , doi =

    2021

  67. [67]

    Wu, Kai and Qin, Xinming and Hu, Wei and Yang, Jinlong , title =. J. Chem. Theory Comput. , volume =. 2022 , doi =

    2022

  68. [68]

    Rettig, Adam and Lee, Joonho and Head-Gordon, Martin , title =. J. Chem. Theory Comput. , volume =. 2023 , doi =

    2023

  69. [69]

    , title =

    Yeh, Chia-Nan and Morales, Miguel A. , title =. J. Chem. Theory Comput. , volume =. 2023 , doi =

    2023

  70. [70]

    , title =

    Yeh, Chia-Nan and Morales, Miguel A. , title =. J. Chem. Theory Comput. , volume =. 2024 , doi =

    2024

  71. [71]

    Luo, Zhaolong and Qin, Xinming and Hu, Wei and Yang, Jinlong , title =. J. Chem. Theory Comput. , volume =. 2025 , doi =

    2025

  72. [72]

    and Hohenstein, Edward G

    Parrish, Robert M. and Hohenstein, Edward G. and Mart. Tensor hypercontraction. J. Chem. Phys. , volume =. 2012 , doi =

    2012

  73. [73]

    , title =

    Matthews, Devin A. , title =. J. Chem. Theory Comput. , volume =. 2020 , doi =

    2020

  74. [74]

    2026 , eprint =

    Yang, Junjie and Zhang, Ning and Yuan, Shunyue and Yu, Jincheng and Ye, Hong-Zhou and Chan, Garnet Kin-Lic , title =. 2026 , eprint =

    2026