Pith. sign in

REVIEW 2 major objections 5 minor 43 references

An analytical Pauli formula and multistage refinement let tensor networks find the first 32 eigenstates of a million-by-million Laplacian on 20 qubits with fidelity above 0.95.

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 · grok-4.5

2026-07-10 12:00 UTC pith:FMUV2B5Z

load-bearing objection Solid methods paper: closed-form Pauli matrix elements via binary encoding plus a multistage ITE initializer that actually cuts Trotter steps on smooth Sturm–Liouville modes. the 2 major comments →

arxiv 2607.08166 v1 pith:FMUV2B5Z submitted 2026-07-09 quant-ph

Efficient Pauli-decomposition and multistage state-refinement for tensor network based differential equation solver

Vishwabhushan Suresh Gholap , Himadri Shekhar Dhar , Siddhartha Santra This is my paper
classification quant-ph
keywords tensor networksmatrix product operatorsPauli decompositionimaginary time evolutionSturm–Liouvillemultistage refinementdifferential operators
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.

Discretizing a differential equation on a fine grid produces huge matrices whose eigenvalues and eigenfunctions are expensive to obtain by classical linear algebra. For Sturm–Liouville operators the authors show that those matrices can be written as matrix product operators on only n = log2(N) qubits, provided two bottlenecks are removed. First they derive a closed-form expression for every Pauli-string coefficient that uses only the binary structure and value numbers of the string, cutting memory from exponential in n to linear. Second they evolve the system first on a coarse grid, linearly interpolate the resulting low-lying states onto a finer grid, and finish the imaginary-time evolution there; because those states are smooth, the interpolation starts the fine-grid calculation near the target and cuts the number of Trotter steps by up to two orders of magnitude. The combined pipeline recovers the first 32 eigenstates of a Laplacian larger than 10^6 with fidelity greater than 0.95, works for a two-dimensional anharmonic oscillator, and fails gracefully only when disorder becomes strong enough to destroy smoothness.

Core claim

For sparse, structured differential operators that become tridiagonal after discretization, an analytic formula for Pauli-string matrix elements (built from two n-bit binary numbers that encode column location and phase) yields every expansion coefficient with O(2n) memory; the resulting compact MPO, together with a coarse-to-fine state-refinement map, makes imaginary-time evolution converge fast enough to extract dozens of accurate low-lying eigenpairs on grids that exceed a million points.

What carries the argument

The Structure and Value Theorems: every Pauli string’s sole nonzero entry in row j sits at column α ⊕ j and equals (−i)^K(α∧β) (−1)^Π(β∧j), where α and β are the binary structure and value numbers of the string. This identity turns the Pauli decomposition of any matrix with known entries into a direct sum of length 2^n, and for tridiagonal operators collapses further to (n+1)2^n or fewer terms.

Load-bearing premise

The low-lying eigenfunctions must be smooth enough that linear interpolation of a coarse-grid solution already lies close to the true fine-grid solution; otherwise the multistage speedup and the reported fidelities disappear.

What would settle it

Run the same 20-qubit pipeline on a Laplacian plus a random potential whose strength is large enough that the ground-state fidelity falls below 0.95 (or the relative energy error exceeds the paper’s 10^{-4} threshold) while keeping bond dimension fixed; if the multistage refinement still converges in far fewer Trotter steps than direct evolution, the smoothness premise is false.

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

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

2 major / 5 minor

Summary. The paper develops two techniques to improve tensor-network (TN) solvers for Sturm–Liouville problems arising from PDEs. First, it derives closed-form expressions (Theorems 1–2, Eq. (2)) for the matrix elements of arbitrary n-qubit Pauli strings via binary structure number α and value number β; this yields Pauli coefficients (Eq. (3)) by a direct sum, reducing memory from O(2^{n+1}) (explicit Pauli matrices) to O(2n) and enabling efficient MPO construction for sparse structured operators such as tridiagonal discretizations (Props. 4–7, specialized sums in Appendix C). Second, it introduces a multistage state-refinement heuristic (linear interpolation map Û : H_n → H_{n+n′}, Eq. (11)) that seeds fine-grid imaginary-time evolution (ITE) from coarse-grid eigenstates, cutting the number of Trotter steps by up to two orders of magnitude. The framework is demonstrated on the Dirichlet Laplacian (first 32 eigenstates of dimension >10^6 at fidelity ≥0.95 with a 20-qubit MPO), the separable 2-D anharmonic oscillator (fidelity >0.99 vs exact diagonalization), and 1-D systems with quadratic/quartic/random potentials (accuracy degrades with disorder strength).

Significance. If the claims hold, the work supplies two reusable algorithmic primitives of independent interest: an analytical Pauli-string matrix-element formula that removes a memory bottleneck for any n-qubit operator (not only differential ones), and a simple coarse-to-fine MPS refinement that accelerates ITE for smooth low-lying modes. The induction proofs (Theorems 1–2, Props. 1–7) are machine-checkable in principle, the numerical benchmarks are external (closed-form Laplacian eigenfunctions, exact diagonalization), and the paper itself maps the failure mode (strong random potentials increase entanglement, rendering low-χ MPOs inaccurate). These contributions are useful both for classical TN PDE solvers and as classical simulators/preconditioners for quantum algorithms that rely on Pauli decompositions or imaginary-time evolution.

major comments (2)
  1. [§II.B, Eq. (3) and abstract] §II.B–C and abstract: The memory claim “O(2^{n+1}) → O(2n)” is correct for the evaluation of each Pauli coefficient once the operator matrix elements are known analytically, but for a completely general dense operator the sum in Eq. (3) still requires O(2^n) storage of those elements. The paper should state more sharply that the O(2n) figure applies only when the operator itself admits an O(n)- or O(poly(n))-space description (as the tridiagonal SL operators do via the functions a_i, b_i, c_i). Without that qualification the abstract overstates the generality.
  2. [§II.E, Eq. (11)] §II.E and Fig. 4: The multistage map Û is asserted to be representable as an MPO so that the whole pipeline stays inside the TN framework for large n. No explicit MPO construction, bond-dimension bound, or numerical verification of the MPO–MPS contraction cost is supplied. For the system sizes shown classical interpolation is used; a short appendix giving the MPO tensors of Û (or a reference to a standard linear-interpolation MPO) would make the scalability claim fully rigorous.
minor comments (5)
  1. [Fig. 2] Fig. 2 caption and main text: the comparison libraries “qiskit” and “pauli_lcu” are cited only by name; full package versions and the precise API calls used would improve reproducibility.
  2. [Tables I–II] Table I vs Table II: wall-clock times are reported on the same hardware, yet the two tables measure different stages (Pauli expansion alone vs full MPO+Trotter). A single combined timing table for the end-to-end pipeline would clarify the relative cost of each component.
  3. [§III.D] §III.D and Fig. 9: truncation-error analysis is performed only up to n=10 because of the exponential growth of Pauli strings. A brief remark on how the observed error scales (or an extrapolation) would help readers judge reliability at the n=20 regime used for the main Laplacian results.
  4. [§III.A] Notation: the same symbol χ is used both for bond dimension and (in places) for the diffusion constant; a distinct symbol for the latter would avoid momentary confusion.
  5. [Appendix C] Appendix C pseudocode: the functions C0/C1/C2 return un-normalized sums; the final division by 2^n and the phase factor appear only in the outer procedure. Adding a one-line comment that the returned values must still be scaled would prevent implementation errors.

Circularity Check

0 steps flagged

No significant circularity: Pauli-coefficient formulas are derived by induction from binary encodings, multistage refinement is an explicit heuristic validated on external exact solutions, and no fitted inputs are renamed as predictions.

full rationale

The load-bearing claims rest on Theorems 1–2 (structure and value numbers for Pauli matrix elements, proved by induction on n) and the resulting closed-form coefficient sum (Eq. 3), which reduce memory for any n-qubit matrix without assuming the target eigenstates or fidelities. Specialization to tridiagonal Sturm–Liouville operators (Props. 4–7) follows directly from the sparsity pattern of the finite-difference stencil and is not defined in terms of the later numerical results. Multistage refinement (§II.E) is presented as a heuristic map Ũ that linearly interpolates coarse-grid MPS amplitudes; its justification is the smoothness of low-lying modes, which the paper itself falsifies for strong random potentials (§III.C) rather than assuming it circularly. All reported fidelities and energy errors are measured against independent external references: closed-form Dirichlet Laplacian eigenfunctions (Fig. 3) and exact diagonalization of coarsened anharmonic/random Hamiltonians (Figs. 6–8). Algorithmic parameters (χ, Δt, truncation thresholds) are free choices whose effect is quantified, not tuned to force agreement with a target formula. Background citations to TN/ITE literature are not load-bearing uniqueness theorems or self-citations that close the derivation. Consequently the derivation chain does not reduce to its own inputs by construction.

Axiom & Free-Parameter Ledger

4 free parameters · 5 axioms · 2 invented entities

The paper rests on standard linear algebra and tensor-network facts plus the modeling choice that the target operators are sparse, structured, and have smooth low modes. Free parameters are algorithmic (bond dimension, Trotter step, truncation, refinement schedule), not physical constants fitted to data. No new physical entities are postulated.

free parameters (4)
  • bond dimension χ (and truncation threshold) = typically 20–50 in reported runs
    Chosen by hand (χ≈10–50 in figures); controls accuracy vs cost of HMPO/EMPO and MPS. Central fidelity claims depend on adequate χ.
  • Trotter step size Δt = 0.1
    Fixed at 0.1 in Fig. 4 and related runs; not systematically optimized. Affects number of steps and Trotter error.
  • multistage refinement schedule (n sequence) = e.g. 4→6→10
    Paths such as 4→6→10 vs direct n=10 are chosen heuristically; speedup claims depend on this schedule.
  • potential strength g (disorder scans) = 2–20 eV in figures
    Scanned parameter in §III.C; not fitted, but marks the regime boundary where the method fails.
axioms (5)
  • standard math n-qubit Pauli strings form a complete orthogonal basis under the Hilbert–Schmidt product Tr(P†Q)=2^n δ_PQ
    Used throughout §II.B and Appendix A to extract coefficients c_ℓ.
  • domain assumption Centered finite-difference discretization of a Sturm–Liouville operator yields a real symmetric tridiagonal matrix with the stated a_i, b_i, c_i
    Proposition 3 / §II.C; standard numerical analysis assumption for the operator class treated.
  • domain assumption Low-lying eigenstates of the considered differential operators are smooth enough that linear interpolation between coarse-grid amplitudes is a good fine-grid initializer
    Stated in §II.E; load-bearing for the multistage speedup claim.
  • domain assumption MPO bond truncation with relative-norm threshold keeps operator error small enough for the reported eigenstate fidelities
    §II.D and §III.D; truncation errors reported <10^{-4} for HMPO on considered sizes.
  • standard math First-order Suzuki–Trotter with commuting Pauli groups is an adequate evolution operator for the ITE runs shown
    §II.D, Eq. 10; standard approximation with controllable step-size error.
invented entities (2)
  • structure number α(P_n) and value number β(P_n) for Pauli strings independent evidence
    purpose: Encode column location and nonzero entry value of any Pauli string so matrix elements can be written in closed form (Eq. 2).
    Binary encodings defined in Defs 1–2; standard idea in stabilizer literature but specialized here into the Structure/Value theorems used for coefficient evaluation.
  • multistage refinement map Û : H_n → H_{n+n′} no independent evidence
    purpose: Linear (MPO-representable) interpolation that lifts coarse-grid eigenstates to fine-grid initial states for ITE.
    Defined in §II.E, Eq. 11–12; algorithmic construct, not a physical object.

pith-pipeline@v1.1.0-grok45 · 26877 in / 3551 out tokens · 40995 ms · 2026-07-10T12:00:37.263379+00:00 · methodology

0 comments
read the original abstract

Classical numerical techniques for solving partial differential equations (PDEs) become computationally expensive as the dimension of the discretized differential operator increases. For PDEs giving rise to Sturm--Liouville problems, tensor network (TN) methods can be highly productive: an operator of dimension $N\times N$ can be represented as a matrix product operator (MPO) using only $n=\log_2(N)$ qubits, enabling computation of eigenvalues and eigenvectors via imaginary time evolution (ITE). However, this remains computationally challenging. First, most methods for generating MPOs of large operators without explicit tensor-product structure require prohibitively large memory. Second, the number of Trotterization steps for convergence in conventional ITE increases rapidly with $n$. We present techniques to mitigate both challenges for certain sparse, structured differential operators. To address the first, we construct the MPO by expanding the operator in the Pauli-string basis, enabled by an analytical expression for the Pauli basis coefficients that reduces the memory requirement from $\mathcal{O}(2^{n+1})$ to $\mathcal{O}(2n)$. To address the second, we propose a multistage state-refinement heuristic that accelerates ITE convergence, reducing convergence time by up to two orders of magnitude. Using this TN framework, we compute the first 32 eigenstates of a Laplacian of dimension exceeding $10^6$ with fidelity above $0.95$ using a 20-qubit MPO. We further validate the method on the 2D anharmonic oscillator and investigate disordered systems, where increasing random potential strength degrades accuracy and limits the approach.

Figures

Figures reproduced from arXiv: 2607.08166 by Himadri Shekhar Dhar, Siddhartha Santra, Vishwabhushan Suresh Gholap.

Figure 1
Figure 1. Figure 1: FIG. 1. Successive SVD truncations used to obtain the MPS [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2. Wall-clock time scaling of the Pauli-basis expansion [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 4
Figure 4. Figure 4: FIG. 4. Fidelity vs. cumulative number of Trotter step for the [PITH_FULL_IMAGE:figures/full_fig_p008_4.png] view at source ↗
Figure 3
Figure 3. Figure 3: FIG. 3. Fidelity of first 50 eigenstate of the Dirichlet Lapla [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figure 5
Figure 5. Figure 5: FIG. 5. Probability density [PITH_FULL_IMAGE:figures/full_fig_p009_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: FIG. 6. Relative eigenvalue error and fidelity of the first [PITH_FULL_IMAGE:figures/full_fig_p009_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: shows the fidelity error and relative error in energy eigenvalues, when bond dimension 𝜒 is varied, for the three different potentials. The strength parameter is set to 𝑔 = 2 eV (solid line) and 𝑔 = 10 eV (dashed line) [PITH_FULL_IMAGE:figures/full_fig_p010_7.png] view at source ↗
Figure 9
Figure 9. Figure 9: FIG. 9. Accumulated bond truncation error in the Hamilto [PITH_FULL_IMAGE:figures/full_fig_p010_9.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

43 extracted references · 43 canonical work pages · 5 internal anchors

  1. [1]

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

    work page 1992

  2. [2]

    S. R. White, Density-matrix algorithms for quantum renormalization groups, Phys. Rev. B48, 10345 (1993)

    work page 1993

  3. [3]

    Schollw¨ ock, The density-matrix renormalization group in the age of matrix product states, Annals of Physics 326, 96 (2011), january 2011 Special Issue

    U. Schollw¨ ock, The density-matrix renormalization group in the age of matrix product states, Annals of Physics 326, 96 (2011), january 2011 Special Issue

    work page 2011

  4. [4]

    Or´ us, A practical introduction to tensor networks: Matrix product states and projected entangled pair states, Annals of Physics349, 117 (2014)

    R. Or´ us, A practical introduction to tensor networks: Matrix product states and projected entangled pair states, Annals of Physics349, 117 (2014)

    work page 2014

  5. [5]

    Verstraete, V

    F. Verstraete, V. Murg, and J. Cirac, Matrix prod- uct states, projected entangled pair states, and vari- ational renormalization group methods for quantum spin systems, Advances in Physics57, 143 (2008), https://doi.org/10.1080/14789940801912366

    work page doi:10.1080/14789940801912366 2008

  6. [6]

    J. I. Cirac, D. P´ erez-Garc´ ıa, N. Schuch, and F. Ver- straete, Matrix product states and projected entangled pair states: Concepts, symmetries, theorems, Rev. Mod. Phys.93, 045003 (2021)

    work page 2021

  7. [7]

    Or´ us, Tensor networks for complex quantum systems, Nature Reviews Physics1, 538 (2019)

    R. Or´ us, Tensor networks for complex quantum systems, Nature Reviews Physics1, 538 (2019)

    work page 2019

  8. [8]

    M. D. Garc´ ıa and A. M´ arquez Romero, Survey on com- putational applications of tensor-network simulations, IEEE Access12, 193212 (2024)

    work page 2024

  9. [9]

    Eisert, M

    J. Eisert, M. Cramer, and M. B. Plenio, Colloquium: Area laws for the entanglement entropy, Rev. Mod. Phys. 82, 277 (2010)

    work page 2010

  10. [10]

    M. B. Hastings, An area law for one-dimensional quan- tum systems, Journal of Statistical Mechanics: Theory and Experiment2007, P08024 (2007)

    work page 2007

  11. [11]

    Vidal, Efficient classical simulation of slightly entan- gled quantum computations, Phys

    G. Vidal, Efficient classical simulation of slightly entan- gled quantum computations, Phys. Rev. Lett.91, 147902 (2003)

    work page 2003

  12. [12]

    Vidal, Efficient simulation of one-dimensional quan- tum many-body systems, Phys

    G. Vidal, Efficient simulation of one-dimensional quan- tum many-body systems, Phys. Rev. Lett.93, 040502 (2004)

    work page 2004

  13. [13]

    P. C. S. Costa, S. Jordan, and A. Ostrander, Quantum algorithm for simulating the wave equation, Phys. Rev. A99, 012323 (2019)

    work page 2019

  14. [14]

    S. Jin, N. Liu, and Y. Yu, Quantum simulation of partial differential equations: Applications and detailed analy- sis, Phys. Rev. A108, 032603 (2023)

    work page 2023

  15. [15]

    I. V. Oseledets, Tensor-train decomposition, SIAM Journal on Scientific Computing33, 2295 (2011), https://doi.org/10.1137/090752286

    work page doi:10.1137/090752286 2011

  16. [16]

    I. V. Oseledets and S. V. Dolgov, Solution of linear systems and matrix inversion in the tt-format, SIAM Journal on Scientific Computing34, A2718 (2012), https://doi.org/10.1137/110833142

    work page doi:10.1137/110833142 2012

  17. [17]

    B. N. Khoromskij, Tensor numerical methods for high- dimensional pdes: Basic theory and initial applications (2014), arXiv:1408.4053 [math.NA]

    work page internal anchor Pith review Pith/arXiv arXiv 2014

  18. [18]

    Kazeev, O

    V. Kazeev, O. Reichmann, and C. Schwab, Low-rank ten- sor structure of linear diffusion operators in the tt and qtt formats, Linear Algebra and its Applications438, 4204 (2013)

    work page 2013

  19. [19]

    S. V. Dolgov and D. V. Savostyanov, Alternating minimal energy methods for linear systems in higher dimensions, SIAM Journal on Scientific Computing36, A2248 (2014), https://doi.org/10.1137/140953289

    work page doi:10.1137/140953289 2014

  20. [20]

    Dolgov, B

    S. Dolgov, B. Khoromskij, I. Oseledets, and D. Savostyanov, Computation of extreme eigenval- ues in higher dimensions using block tensor train format, Computer Physics Communications185, 1207 (2014)

    work page 2014

  21. [21]

    Holtz, T

    S. Holtz, T. Rohwedder, and R. Schneider, The alternat- ing linear scheme for tensor optimization in the tensor train format, SIAM Journal on Scientific Computing34, A683 (2012), https://doi.org/10.1137/100818893

    work page doi:10.1137/100818893 2012

  22. [22]

    A. M. Boelens, D. Venturi, and D. M. Tartakovsky, Paral- lel tensor methods for high-dimensional linear pdes, Jour- nal of Computational Physics375, 519 (2018)

    work page 2018

  23. [23]

    A. J. Daley, C. Kollath, U. Schollw¨ ock, and G. Vidal, Time-dependent density-matrix renormalization-group using adaptive effective hilbert spaces, Journal of Statis- tical Mechanics: Theory and Experiment2004, P04005 (2004)

    work page 2004

  24. [24]

    Vidal, Classical simulation of infinite-size quantum lattice systems in one spatial dimension, Phys

    G. Vidal, Classical simulation of infinite-size quantum lattice systems in one spatial dimension, Phys. Rev. Lett. 98, 070201 (2007)

    work page 2007

  25. [25]

    Haegeman, J

    J. Haegeman, J. I. Cirac, T. J. Osborne, I. Piˇ zorn, H. Ver- schelde, and F. Verstraete, Time-dependent variational principle for quantum lattices, Phys. Rev. Lett.107, 070601 (2011)

    work page 2011

  26. [26]

    Paeckel, T

    S. Paeckel, T. K¨ ohler, A. Swoboda, S. R. Manmana, U. Schollw¨ ock, and C. Hubig, Time-evolution methods for matrix-product states, Annals of Physics411, 167998 (2019)

    work page 2019

  27. [27]

    A tree-approach Pauli decomposition algorithm with application to quantum computing

    O. Koska, M. Baboulin, and A. Gazda, A tree-approach pauli decomposition algorithm with application to quan- tum computing (2024), arXiv:2403.11644 [quant-ph]

    work page internal anchor Pith review Pith/arXiv arXiv 2024

  28. [28]

    Stabilizer Codes and Quantum Error Correction

    D. Gottesman, Stabilizer codes and quantum error cor- rection (1997), arXiv:quant-ph/9705052 [quant-ph]

    work page internal anchor Pith review Pith/arXiv arXiv 1997

  29. [29]

    A. W. Harrow, A. Hassidim, and S. Lloyd, Quantum al- gorithm for linear systems of equations, Phys. Rev. Lett. 103, 150502 (2009)

    work page 2009

  30. [30]

    A. M. Childs, R. Kothari, and R. D. Somma, Quantum algorithm for systems of linear equations with exponentially improved dependence on preci- sion, SIAM Journal on Computing46, 1920 (2017), https://doi.org/10.1137/16M1087072

    work page internal anchor Pith review doi:10.1137/16m1087072 1920

  31. [31]

    D. W. Berry, High-order quantum algorithm for solving linear differential equations, Journal of Physics A: Math- ematical and Theoretical47, 105301 (2014)

    work page 2014

  32. [32]

    A. M. Childs, J.-P. Liu, and A. Ostrander, High-precision quantum algorithms for partial differential equations, Quantum5, 574 (2021)

    work page 2021

  33. [33]

    Farghadan, M

    A. Farghadan, M. M. Masteri Farahani, and M. Akbari, A fast quantum algorithm for solving partial differential equations, Scientific Reports15, 5317 (2025)

    work page 2025

  34. [34]

    Arseniev, D

    B. Arseniev, D. Guskov, R. Sengupta, J. Biamonte, and I. Zacharov, Tridiagonal matrix decomposition for hamil- tonian simulation on a quantum computer, Phys. Rev. A 109, 052629 (2024)

    work page 2024

  35. [35]

    Peruzzo, J

    A. Peruzzo, J. McClean, P. Shadbolt, M.-H. Yung, X.-Q. Zhou, P. J. Love, A. Aspuru-Guzik, and J. L. O’Brien, A variational eigenvalue solver on a photonic quantum processor, Nature Communications5, 4213 (2014)

    work page 2014

  36. [36]

    J. R. McClean, J. Romero, R. Babbush, and A. Aspuru- Guzik, The theory of variational hybrid quantum- classical algorithms, New Journal of Physics18, 023023 (2016)

    work page 2016

  37. [37]

    Cerezo, A

    M. Cerezo, A. Arrasmith, R. Babbush, S. C. Benjamin, 13 S. Endo, K. Fujii, J. R. McClean, K. Mitarai, X. Yuan, L. Cincio, and P. J. Coles, Variational quantum algo- rithms, Nature Reviews Physics3, 625 (2021)

    work page 2021

  38. [38]

    Lubasch, J

    M. Lubasch, J. Joo, P. Moinier, M. Kiffner, and D. Jaksch, Variational quantum algorithms for nonlinear problems, Phys. Rev. A101, 010301 (2020)

    work page 2020

  39. [39]

    McArdle, T

    S. McArdle, T. Jones, S. Endo, Y. Li, S. C. Benjamin, and X. Yuan, Variational ansatz-based quantum simula- tion of imaginary time evolution, npj Quantum Informa- tion5, 75 (2019)

    work page 2019

  40. [40]

    Motta, C

    M. Motta, C. Sun, A. T. K. Tan, M. J. O’Rourke, E. Ye, A. J. Minnich, F. G. S. L. Brand˜ ao, and G. K.-L. Chan, Determining eigenstates and thermal states on a quan- tum computer using quantum imaginary time evolution, Nature Physics16, 205 (2020)

    work page 2020

  41. [41]

    Bachmayr, Low-rank tensor methods for partial dif- ferential equations, Acta Numerica32, 1–121 (2023)

    M. Bachmayr, Low-rank tensor methods for partial dif- ferential equations, Acta Numerica32, 1–121 (2023)

    work page 2023

  42. [42]

    T. N. Georges, B. K. Berntson, C. S¨ underhauf, and A. V. Ivanov, Pauli decomposition via the fast walsh-hadamard transform (2025), arXiv:2408.06206 [quant-ph]

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  43. [43]

    M. Suzuki, Generalized Trotter’s formula and systematic approximants of exponential operators and inner deriva- tions with applications to many-body problems, Commu- nications in Mathematical Physics51, 183 (1976). 14 Appendix A: Definitions and Notations The single-qubit Pauli matrices, P∞ = 1 0 0 1 , 1 0 0−1 , 0 1 1 0 , 0−𝑖 𝑖0 ≡ {𝐼, 𝑍, 𝑋, 𝑌}, span the o...

    work page 1976