Pith. sign in

REVIEW 2 major objections 5 minor 1 cited by

Operator-centric Clifford algebra for variational eigensolvers and finite-shot adaptive selection

T0 review · 2 major / 5 minor · reviewed 2026-08-03 · deepseek-v4-flash

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

Pith's one-line read For real Hamiltonians and real states, every even-Y Pauli word has zero ADAPT gradient, and finite-shot racing beats fixed selection 84/100 to 0/100.

desk verdict Solid algebra and honest about its limits, but the 0/100 vs 84/100 finite-shot comparison conflates adaptivity with shot budget. read the letter →

arxiv 2607.17443 v2 pith:PELY3Q6R submitted 2026-07-20 quant-ph

Ginanjar Utama , Hermawan Kresno Dipojono This is my paper
classification quant-ph MSC 15A6681P68 PACS 03.67.-a
keywords CliffordalgebraADAPT-VQEtransposeparityPauliwordsvariationalquantumeigensolverfinite-shotselectionconfidence-boundracingtransverse-fieldIsingmodel
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

This paper establishes a precise algebraic foundation for variational quantum eigensolvers by representing states, gates, observables, and channels as sparse elements of the complex Clifford algebra, and it proves an exact selection rule for real-state ADAPT-VQE. For any real Hamiltonian and real state, every candidate Pauli word with an even number of Y factors has exactly zero gradient, while odd-Y rotations keep the state real; the ADAPT pool can therefore be pruned by a one-line parity test. On the measurement side, 100-seed finite-shot simulations show that a single fixed batch of shots fails every time at n=4, while cumulative escalation and confidence-bound racing each succeed in 84/100 runs, with racing using one-third fewer median shots. A systematic contiguous three-local odd-Y pool makes exact ADAPT accurate to below 1.3e-12 for n up to 6 at the critical transverse-field Ising point. The contribution is a corrected algebraic dictionary, a proof of the real-sector filter, and a reproducible finite-shot baseline.

What carries the argument

The load-bearing object is the transpose-parity rule on Pauli words: the count of Y factors, νY(W), determines whether a word is real symmetric (even Y) or purely imaginary antisymmetric (odd Y). For real H and real ρ, Proposition 1 gives Tr(ρ[H,P])=0 for even-Y P, so such words can be dropped from the ADAPT pool without approximation; odd-Y rotations preserve realness, making the rule inductive across the whole ansatz build. This is carried inside a sparse operator-centric Clifford algebra in which the Jordan-Wigner strings map anticommuting generators to Pauli words, phase-normalized blades make every blade Hermitian, and one Pauli rotation at most doubles operator support. The finite-shot

What would settle it

Compute Tr(rho[H,P]) for any real symmetric H, any real symmetric rho, and a Pauli word with an even number of Y factors; Proposition 1 predicts exactly zero, so a nonzero value would refute it. Separately, give the fixed policy racing's median total shot budget (about 0.53M shots) and rerun the n=4 100-seed test; if fixed then succeeds at comparable rates, the headline adaptivity claim is a budget artifact.

Watch

Extended reading notes

Core claim

The central claim is that the familiar odd-Y restriction in real-state qubit ADAPT is not a heuristic but an exact transpose-parity statement. With computational-basis transposition acting as W^T = (-1)^{νY(W)} W on Pauli words, a real symmetric Hamiltonian and a real symmetric density matrix make Tr(ρ[H,P]) vanish for every even-Y word, and odd-Y rotations preserve the real sector by induction. The paper builds a sparse operator-centric Clifford algebra Cl(2n,C) ≅ M(2^n,C) using the Jordan-Wigner map to make Pauli words exact phase-normalized blades, and derives gradients for shared-parameter variational ansatzes as sums of physical-gate contributions. Empirically, exact ADAPT with a system

Load-bearing premise

The finite-shot policy ordering rests on pointwise 3-sigma confidence bounds with independent per-word binomial sampling being sufficient to eliminate candidates, and on the fixed baseline receiving far fewer total shots than the adaptive policies.

Editorial extensions

If this is right

  • The even-Y sector ((4^n + 2^n)/2 of the 4^n Pauli words) can be removed from real-state ADAPT pools exactly; only odd-Y words need be measured.
  • At a 4096-shot ceiling and n=4, a fixed 128-shot ranking succeeds in 0/100 seeds, whereas uniform escalation and racing both succeed in 84/100 with 95% Wilson interval [75.6, 89.9].
  • Racing attains the same success as uniform escalation while lowering median total shots from 0.803M to 0.530M, a 34% reduction.
  • In Hamiltonian variational ansatz layers, shifting a shared parameter for the whole layer is not a valid two-point shift even when gates commute; exact gradients must be summed over physical gates.
  • A compact Z-dressed Y pool is exact only at n=4; the contiguous three-local odd-Y pool reaches relative errors below 1.3e-12 through n=6.

Reading between the lines

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

  • The 0/100 versus 84/100 gap is not an equal-cost comparison: fixed gets 128 shots per word once, while escalation and racing are allowed up to 4096 cumulative shots, so part of the measured advantage is simply a larger measurement budget.
  • Equalizing total shots between fixed and racing budgets would be a sharper test of adaptivity; the present data do not rule out a fixed policy that spends racing's median 0.53M shots up front.
  • The racing heuristic's 3-sigma thresholds, winner's bias, and word-independence assumptions are not corrected for multiple comparisons; a covariance-aware variant that measures commuting words together could plausibly reduce the shot count further.
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

2 major / 5 minor

Summary. The paper develops a sparse Pauli-word (operator-centric) realization of n-qubit operators in the complex Clifford algebra Cl(2n,C), and applies it to variational eigensolvers. It gives a phase-normalized blade/Pauli dictionary, distinguishes Pauli-word rotations from Spin rotors, derives a transpose-parity selection rule (Proposition 1) that excludes even-Y Pauli words for real-symmetric Hamiltonians/states, and derives exact shared-parameter HVA gradients (Eq. 23). Numerical work on the critical transverse-field Ising chain includes HVA, exact ADAPT with two pools, and finite-shot comparisons of fixed, uniform-escalation, and confidence-bound-racing selection policies. The paper is explicitly framed as a corrected algebraic formulation and a reproducible finite-shot baseline, not as a speedup claim.

Significance. The algebraic core is rigorous and well validated: dense-matrix checks, correct phase conventions, and a clean proof of the real-sector filter. The explicit admission that odd-Y pools are not new (with prior work cited) is appropriate. The paper also ships a reproducible code/data package, which is a genuine strength. If the finite-shot policy ordering is correct, the 34% shot saving from racing would be a useful baseline. However, the central empirical policy claim is currently undermined by a missing equal-cost control, so the significance is conditional on that being fixed.

major comments (2)
  1. [Sec. V C, Table III] The headline comparison is not equal-cost. Fixed selection measures each candidate once at the base budget of 128 shots, while uniform escalation and racing escalate cumulatively to a 4096-shot per-word ceiling. Median total shots are 18,816 (fixed) vs 802,816 (uniform) and 530,176 (racing). Thus the 0/100 vs 84/100 gap may be due to shot budget rather than adaptivity. The ceiling sweep in Table IV and Fig. 3 demonstrates the dominant role of budget: at base 64, success rises from 4% at ceiling 512 to 89% at ceiling 8192. Without a fixed policy at the same per-word ceiling (e.g., one batch of 4096 shots per candidate) or a total-shot-matched control, the conclusion that 'cumulative escalation is essential' is unsupported. The text in Sec. III C says the policies are compared at a 'common base budget and ceiling,' but fixed never reaches the ceiling; the wording is misleading.
  2. [Sec. III C, Eqs. (31)-(35), Table III and Table IV] The confidence-bound racing rule uses pointwise 3-sigma intervals (kappa=3) with no multiple-comparison correction, and Eq. (31) introduces a pseudocount. These choices directly control the elimination decisions. The observed success rate 84/100 sits between the ceiling-sweep values 49/100 and 89/100, and Table S4 shows high 'ambiguous fractions' (0.88-0.997), indicating frequent rank ambiguity. The paper acknowledges these are heuristics (Sec. VII) but does not quantify sensitivity. A kappa sweep (e.g., 2, 3, 4) and a check of the pseudocount's influence would establish whether the 84/100 success and the 34% shot saving are robust or artifacts of the chosen interval width. This is load-bearing for the finite-shot claim.
minor comments (5)
  1. [Sec. III C, Sec. V C] The phrase 'common base budget and ceiling per measured Pauli word' should be clarified: fixed uses only the base budget, so the comparison is common in base and ceiling but not in total shot expenditure.
  2. [Sec. IV] The deterministic Adam fallback is mentioned but not described; specify when it is triggered and how its hyperparameters are set.
  3. [Eq. (19)] The count NoddY is correct, but a short derivation or reference would help readers verify the formula.
  4. [Sec. II B, Eq. (10)] The phase factor q=k(k-1)/2 is used in the main text but defined only in Appendix A; a forward reference would improve readability.
  5. [Sec. V C, Fig. 2] Panel (b) shows median total shots, but the accompanying text does not mention the IQR values reported in the Supplemental Material; consider adding one sentence noting the overlap/non-overlap of the shot distributions.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the algebraic derivations are proved from stated assumptions, and the finite-shot policy limitations are acknowledged as heuristics rather than recycled as predictions.

full rationale

The paper's central algebraic claims are self-contained. Lemma 1 is proved by explicit construction (odd-Y words give real orthogonal rotations), and Proposition 1 follows directly from transpose parity and the vanishing trace of a symmetric-antisymmetric product; no fitted parameter or prior result is needed. The shared-parameter gradient formula (Eq. 23) is derived from physical-gate parameter shifts and independently checked against finite differences and dense-matrix invariants in Supplemental Sec. S1, so it is not circular. The finite-shot policy comparison uses hand-set heuristics (kappa=3, Wilson intervals, base/ceiling budgets) that are not fitted to the data; Sec. VII explicitly states that the '3σ signal, gap, and racing rules are heuristics without familywise coverage.' The fixed-versus-adaptive comparison is not equal-cost (fixed uses only the 128-shot base while adaptive can escalate to 4096 shots per word), but that is a fairness/correctness limitation, not a circularity: the successful adaptive runs are not constructed to match the 0/100 fixed outcome. The only overlapping self-citation is ref. [10], used as contextual related work on shot allocation and outcome reuse; it is not load-bearing for any theorem or numerical conclusion. There is no self-citation chain, no imported uniqueness theorem, and no ansatz smuggled in via citation. The results are benchmarked against exact diagonalization and dense-matrix verification, so the derivation chain remains independent of its own conclusions.

Assumptions & free parameters 8 free parameters · 6 assumptions · 0 invented entities

The paper introduces no new physical entities or forces. Its free parameters are protocol choices (kappa, budgets, thresholds, pool cap) that affect the empirical results, not fitted constants of a theory. The axioms are standard algebra plus explicit domain restrictions to the real sector and independent-binomial shot model.

free parameters (8)
  • confidence-bound width kappa = 3
    Used in Eq. (33)-(35) for upper/lower bounds and elimination; success rates and shot savings depend on this width, with no sensitivity sweep.
  • base shots per Pauli word = 128 (64 in racing ceiling sweep)
    Starting measurement budget; the 0/100 fixed result is specific to the 128-shot base.
  • shot ceiling per Pauli word = 4096 (512, 2048, 8192 in sweep)
    Caps the cumulative per-word budget; the ceiling controls the finest resolvable gradient gap and largely drives success rate.
  • pseudocount in variance estimator = 1/2
    Inserted in Eq. (31) to avoid zero estimated variance; changes the confidence bounds and hence racing decisions.
  • ADAPT gradient threshold = 1e-7
    Stops operator addition in exact ADAPT; reported residual errors and operator counts depend on this threshold.
  • success threshold for finite-shot runs = relative energy error < 1e-3
    Defines whether a seed is successful; the 0/100 and 84/100 numbers are specific to this accuracy target.
  • local-three pool operator cap = 20 appended operators
    Limits ADAPT runs; the n=6 local3 result hits the cap, so the reported 1.29e-12 residual is not necessarily the converged pool minimum.
  • HVA depth schedule and continuation offset = p = 1..3; new layer initialized to (0.05, 0.05)
    The reported HVA energy errors are protocol-specific; different depth schedules or initialization would give different residuals.
assumptions (6)
  • standard math Clifford algebra isomorphism Cl(2n,C) ~= M(2^n,C) and the existence of an irreducible gamma-matrix representation.
    Invoked in Sec. II A so that every n-qubit operator can be represented as one algebra element rather than a separate spinor and endomorphism algebra.
  • standard math Jordan-Wigner identification gamma_{2j} = Z...Z X_j, gamma_{2j+1} = Z...Z Y_j gives canonical anticommutation.
    Eqs. (7)-(9); the exact bridge between Pauli words and Clifford generators. The authors correctly note a direct local assignment fails for n>1.
  • domain assumption Hamiltonian and reference state are real symmetric in the computational basis.
    The TFIM J=h=1 Hamiltonian (Eq. 1) and the |+...+> initial state are real; Lemma 1 and Proposition 1 require real-symmetric H and rho.
  • ad hoc to paper The chosen odd-Y local pools are expressive enough to reach the ground-state subspace.
    No expressivity proof is given; the compact pool stagnates at n=5,6 and the local-three pool is hand-chosen. This is stated in Sec. V B and the Limitations.
  • domain assumption Finite-shot noise is independent per Pauli word, exactly binomial, and parameter reoptimization is exact.
    Protocol in Secs. IV and V; device decoherence, state-preparation noise, optimizer noise, and covariance between commuting words are excluded, as acknowledged in Sec. VII.
  • ad hoc to paper Pointwise 3-sigma confidence bounds without multiple-comparison correction are sufficient for candidate elimination.
    Eqs. (33)-(35); the paper states these are heuristics without familywise coverage and that winner-selection bias is not corrected.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Operator-centric Clifford algebra for variational eigensolvers and finite-shot adaptive selection." pith.science (2026). https://pith.science/paper/PELY3Q6R

@misc{pith2026260717443,
  author       = {Pith},
  title        = {Pith review of: Operator-centric Clifford algebra for variational eigensolvers and finite-shot adaptive selection},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/PELY3Q6R}},
  note         = {Machine review of arXiv:2607.17443}
}
abstract

We develop a sparse operator-centric realization of $n$-qubit variational quantum algorithms in the complex Clifford algebra $\mathrm{Cl}(2n,\mathbb{C}) \cong M(2^n,\mathbb{C})$. Density operators, gates, observables, channels, fermionic modes, and adaptive-selection observables are represented in one Pauli-word algebra, with the Jordan--Wigner map providing the exact bridge to anticommuting Clifford generators. We distinguish general Pauli-word rotations from Spin-group rotors and formulate the familiar odd-$Y$ restriction for real-state adaptive ansatzes as an exact transpose-parity statement: for real Hamiltonians and real states, every candidate Pauli word containing an even number of $Y$ factors has zero ADAPT gradient, while odd-$Y$ rotations preserve the real sector. For the critical open transverse-field Ising chain, a depth-three Hamiltonian variational ansatz gives relative energy errors $4.84\times10^{-5}$, $2.19\times10^{-3}$, and $3.67\times10^{-3}$ for $n=4,5,6$. A compact local ADAPT pool is exact at $n=4$ but leaves residual errors at larger sizes; a systematic contiguous three-local odd-$Y$ pool reaches relative errors below $1.3\times10^{-12}$ for $n\leq6$. In 100-seed finite-shot tests at $n=4$, fixed-shot selection succeeds in $0/100$ runs, whereas uniform escalation and confidence-bound racing each succeed in $84/100$ runs; racing lowers median shots by $34\%$. We claim no asymptotic speedup over matrix methods. The contribution is a corrected algebraic formulation, a density-operator derivation and implementation of the real-sector pool filter, and a reproducible study of measurement-limited adaptive selection.

Figures

Figures reproduced from arXiv: 2607.17443 by the authors.

Figure 1
Figure 1. FIG. 1. Deterministic critical-point results. (a) Relative energy error of the Hamiltonian variational ansatz versus depth. (b) [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
Figure 1
Figure 1. FIG. 1. Deterministic critical-point results. (a) Relative energy error of the Hamiltonian variational ansatz versus depth. (b) [PITH_FULL_IMAGE:figures/full_fig_p006_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2. Finite-shot policy comparison at [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figures from the paper (2 more)
Figure 3
Figure 3. Figure 3: FIG. 3. Racing with a fixed base of 64 shots and a varying per-word ceiling. (a) Success probability with 95% Wilson intervals. [PITH_FULL_IMAGE:figures/full_fig_p006_3.png]
Figure 3
Figure 3. Figure 3: FIG. 3. Racing with a fixed base of 64 shots and a varying per-word ceiling. (a) Success probability with 95% Wilson intervals. [PITH_FULL_IMAGE:figures/full_fig_p007_3.png]

Review history (2 revisions) →

Discussion (0). Sign in 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. Adaptive operator-generated subspaces for effective many-body Hamiltonians

    quant-ph 2026-08 conditional novelty 6.0 of 10

    A-CASE reconstructs effective many-body Hamiltonians in adaptive operator-generated subspaces, matching exact FCI on H4 and reaching chemical accuracy with a warm-started reference.

Reference graph

Works this paper leans on

36 extracted references · 7 linked inside Pith · cited by 1 Pith paper

  1. [1]

    arXiv:2607.17443v2 [quant-ph] 31 Jul 2026 2

    We give a phase-consistent Pauli-word/Clifford- blade dictionary and distinguish Pauli-word rota- tions, Spin rotors, and the quantum-information Clifford group. arXiv:2607.17443v2 [quant-ph] 31 Jul 2026 2

    arXiv 2026

  2. [2]

    We cast the established odd- Y restriction for real- state qubit-ADAPT as an exact density-operator transpose-parity rule, including an inductive real- sector preservation statement

  3. [3]

    We derive exact gradients for shared-parameter Hamiltonian variational ansatzes as sums of physical-gate contributions, clarifying why a naive two-point shift of an aggregate parameter can fail even when the gates within a layer commute

  4. [4]

    Clifford group

    We compare fixed-shot ranking, uniform cumulative escalation, and confidence-bound racing at a com- mon shot ceiling, using 100 independent seeds and Wilson confidence intervals. We benchmark the framework on the open transverse- field Ising model (TFIM), H(J, h) =−J n−2X j=0 ZjZj+1 −h n−1X j=0 Xj,(1) whose thermodynamic critical point ish/J= 1 [21]. II. ...

  5. [5]

    Fixed: measure every candidate once at the base budget

  6. [6]

    Uniform escalation: double cumulative shots for every remaining candidate until the statistical gates pass or the ceiling is reached

  7. [7]

    This is a confidence-bound successive-elimination policy in the best-arm formulation of generator selection [ 11]

    Racing: use the same cumulative doubling, but remove a candidate i when its upper confidence bound is below the best lower bound. This is a confidence-bound successive-elimination policy in the best-arm formulation of generator selection [ 11]. Withκ= 3, Ui =|bgi|+κbσi, L i = max(0,|bgi| −κbσi),(33) and retainionly ifU i ≥max j Lj. The leading candidate m...

  8. [8]

    Majland, P

    M. Majland, P. Ettenhuber, and N. T. Zinner, Physical Review A108, 052422 (2023)

    2023

Show all 36 references
  1. [9]

    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, Nature Communications5, 4213 (2014)

    2014

  2. [10]

    H. R. Grimsley, S. E. Economou, E. Barnes, and N. J. Mayhall, Nature Communications10, 3007 (2019)

    2019

  3. [11]

    H. L. Tang, V. O. Shkolnikov, G. S. Barron, H. R. Grims- ley, N. J. Mayhall, E. Barnes, and S. E. Economou, PRX Quantum2, 020310 (2021)

    2021

  4. [12]

    Mitarai, M

    K. Mitarai, M. Negoro, M. Kitagawa, and K. Fujii, Phys- ical Review A98, 032309 (2018)

    2018

  5. [13]

    Schuld, V

    M. Schuld, V. Bergholm, C. Gogolin, J. Izaac, and N. Kil- loran, Physical Review A99, 032331 (2019)

    2019

  6. [14]

    Wierichs, J

    D. Wierichs, J. Izaac, C. Wang, and C. Y.-Y. Lin, Quan- tum6, 677 (2022)

    2022

  7. [15]

    P. G. Anastasiou, N. J. Mayhall, E. Barnes, and S. E. Economou, arXiv preprint arXiv:2306.03227 (2023), arXiv:2306.03227 [quant-ph]

    2023 arXiv

  8. [16]

    Cafaro, N

    C. Cafaro, N. Bahreyni, and L. Rossetti, arXiv preprint arXiv:2405.08152 (2024), arXiv:2405.08152 [quant-ph]

    2024 arXiv

  9. [17]

    C. K. Long, K. Dalton, C. H. W. Barnes, D. R. M. Arvidsson-Shukur, and N. Mertig, Physical Review A 109, 042413 (2024)

    2024

  10. [18]

    Ikhtiarudin, G

    A. Ikhtiarudin, G. K. Sunnardianto, F. Fathurrahman, M. K. Agusta, and H. K. Dipojono, arXiv preprint arXiv:2507.16879 (2025), arXiv:2507.16879 [quant-ph]

    2025 arXiv

  11. [19]

    Huang and A

    R. Huang and A. F. Izmaylov, arXiv preprint arXiv:2509.14917 (2025), arXiv:2509.14917 [quant-ph]

    2025

  12. [20]

    Scriva, N

    G. Scriva, N. Astrakhantsev, S. Pilati, and G. Mazzola, Physical Review A109, 032408 (2024)

    2024

  13. [21]

    S. S. Somaroo, D. G. Cory, and T. F. Havel, Physics Letters A240, 1 (1998)

    1998

  14. [22]

    T. F. Havel and C. J. L. Doran, arXiv preprint quant- ph/0004031 (2000), arXiv:quant-ph/0004031

    2000

  15. [23]

    Hrdina, A

    J. Hrdina, A. N´ avrat, and P. Vaˇ s ´ ık, Quantum Information Processing21, 310 (2022)

    2022

  16. [24]

    Aaronson and D

    S. Aaronson and D. Gottesman, Physical Review A70, 052328 (2004)

    2004

  17. [25]

    Silva, A unified clifford algebra framework for quantum computation: From specific spinor realizations to general operator representations (2025), preprint

    L. Silva, A unified clifford algebra framework for quantum computation: From specific spinor realizations to general operator representations (2025), preprint

    2025

  18. [26]

    M¨ uller, A

    L. M¨ uller, A. B¨ arligea, A. Knapp, and J. S. Kottmann, inSoftware Engineering 2026 Workshops (2026) arXiv:2601.02233 [quant-ph]

    2026 arXiv

  19. [27]

    Kr¨ otz and D

    F. Kr¨ otz and D. Kranzlm¨ uller, arXiv preprint arXiv:2605.25974 (2026), arXiv:2605.25974 [quant-ph]

    2026 arXiv

  20. [28]

    X. Li, Z. Sun, Y. Fan, J. Liu, Z. Li, and J. Yang, arXiv preprint arXiv:2606.28952 (2026), arXiv:2606.28952 [quant-ph]

    2026 arXiv

  21. [29]

    Pfeuty, Annals of Physics57, 79 (1970)

    P. Pfeuty, Annals of Physics57, 79 (1970). 9

    1970

  22. [30]

    Lounesto,Clifford Algebras and Spinors, 2nd ed

    P. Lounesto,Clifford Algebras and Spinors, 2nd ed. (Cam- bridge University Press, Cambridge, 2001)

    2001

  23. [31]

    Jordan and E

    P. Jordan and E. Wigner, Zeitschrift f¨ ur Physik47, 631 (1928)

    1928

  24. [33]

    Tolar, Journal of Physics: Conference Series1071, 012022 (2018)

    J. Tolar, Journal of Physics: Conference Series1071, 012022 (2018)

    2018

  25. [34]

    Mukherjee, N

    A. Mukherjee, N. F. Berthusen, J. C. Getelina, P. P. Orth, and Y.-X. Yao, Communications Physics6, 4 (2023)

    2023

  26. [35]

    E. B. Wilson, Journal of the American Statistical Associ- ation22, 209 (1927)

    1927

  27. [36]

    Operator-centric Clifford algebra for variational eigensolvers and finite-shot adaptive selection

    H.-Y. Huang, R. Kueng, and J. Preskill, Nature Physics 16, 1050 (2020). Supplemental Material for “Operator-centric Clifford algebra for variational eigensolvers and finite-shot adaptive selection” Ginanjar Utama and Hermawan Kresno Dipojono S1. Verification matrix The release...

    2020

  28. [4096]

    At n = 4, the seventh local-three angle is below 3 × 10−8, so the numerically relevant circuit is effectively six-parameter, consistent with the compact result

    Table II summarizes the comparison. At n = 4, the seventh local-three angle is below 3 × 10−8, so the numerically relevant circuit is effectively six-parameter, consistent with the compact result. C. Finite-shot selection policies Figure 2 and Table III compare the three polic...

Pith tools

Reviewed August 3, 2026 · model on record in the stance chip above.