Pith. sign in

REVIEW 3 major objections 6 minor 69 references

Local unitary basis rotations that minimize a cheap non-stoquasticity cost function improve the Quantum Monte Carlo average sign on frustrated Heisenberg magnets.

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

2026-07-31 08:03 UTC pith:6HMB4XBE

load-bearing objection Solid, usable advance on sign-optimized SSE: unitary non-stoquasticity works in practice on ladders and maple-leaf, even if the gradient-alignment story is thinner than the abstract implies. the 3 major comments →

arxiv 2607.24679 v1 pith:6HMB4XBE submitted 2026-07-27 cond-mat.str-el cond-mat.stat-mech

Sign-optimized Quantum Monte Carlo

Julius S. Herz , Robin Sch\"afer , Mat\'ias G. Gonzalez , David J. Luitz This is my paper
classification cond-mat.str-el cond-mat.stat-mech
keywords sign problemquantum Monte Carlostochastic series expansionbasis optimizationnon-stoquasticityfrustrated magnetsmaple-leaf latticeHeisenberg antiferromagnet
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 sign problem ruins the efficiency of Quantum Monte Carlo for frustrated magnets. This paper shows that one can systematically lessen it by rotating the local basis of small clusters so that the phases of the off-diagonal matrix elements of the bond Hamiltonian are minimized. The quantity they minimize, a generalized non-stoquasticity, is cheap to evaluate and needs no Monte Carlo sampling; its local minima turn out to coincide with local maxima of the average sign itself. On fully-frustrated and triangular ladders the procedure recovers known sign-free bases when they exist and finds better bases when they do not. On the two-dimensional maple-leaf lattice it maps a computational phase diagram in which optimized two- and three-spin bases outperform both the ordinary computational basis and the usual dimer/trimer eigenbases over large regions of coupling space. The resulting sign improvement lets the authors reach lower temperatures, matching the reach of numerical linked-cluster expansions on the same model.

Core claim

Local unitary rotations that minimize the generalized non-stoquasticity of the bond Hamiltonian produce simulation bases whose average Quantum Monte Carlo sign is at least as good as, and in large parts of parameter space strictly better than, the computational basis and the corresponding cluster eigenbases; the local minima of that cost function coincide with local maxima of the average sign for the models studied.

What carries the argument

Generalized non-stoquasticity S(U): the sum over off-diagonal bond-Hamiltonian matrix elements of half the difference between their modulus and their real part. Minimizing S(U) over local unitary rotations on two- or three-spin clusters supplies the candidate bases.

Load-bearing premise

That the cheap non-stoquasticity cost and the true average sign share the same local optima, and that their gradients stay aligned enough for ordinary first-order optimizers to improve the sign—an alignment shown only numerically on the lattices considered, not proved in general.

What would settle it

On any of the studied models, locate a local minimum of S(U) whose average sign is worse than that of a nearby basis, or run the same optimizer and find that the sign does not rise when S falls; either observation would break the claimed coincidence of extrema.

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

If this is right

  • Optimized local bases extend the lowest usable temperature of SSE-QMC by factors of two to four relative to the computational or cluster-eigenbasis on the maple-leaf and triangular-ladder models.
  • A computational phase diagram can be drawn that shows, point by point, which clustering and which unitary class yields the best sign.
  • When the effective superlattice is bipartite the same optimizer recovers a fully stoquastic (sign-free) representation, including the Marshall-sign case.
  • Direct thermodynamic comparisons with numerical linked-cluster expansions become possible at temperatures previously inaccessible to sign-plagued QMC.

Where Pith is reading between the lines

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

  • The same cost-function machinery should transfer immediately to other frustrated lattices whose natural clusters are small (breathing kagome, pyrochlore tetrahedra, Shastry–Sutherland dimers).
  • Because S is noise-free and system-size independent, it can be used as a pre-processing filter to decide whether a given clustering is worth the full Monte Carlo investment.
  • If the coincidence of extrema continues to hold for larger clusters, tensor-network or machine-learned unitaries on bigger blocks become a natural next step.

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 / 6 minor

Summary. The manuscript proposes to alleviate the QMC sign problem by optimizing local unitary basis rotations on small clusters, using as a cost function a "generalized non-stoquasticity" S(U) (Eq. 8) — a noise-free, QMC-independent quantity that vanishes iff the bond Hamiltonian is stoquastic. The authors claim (i) that minimizing S(U) with Adam+LBFGS from Haar-random starts yields bases whose measured average sign ⟨s⟩ exceeds that of the computational and cluster eigenbases over large regions of parameter space, and (ii) that local minima of S coincide with local maxima of ⟨s⟩ and the gradients of the two quantities are aligned during optimization. The method is validated on the fully frustrated ladder (where it recovers the known sign-free rung basis exactly, S=0), applied to the triangular ladder (where genuinely complex non-orthogonal optima outperform both reference bases), and then used to map a "computational phase diagram" of the anisotropic maple-leaf lattice Heisenberg antiferromagnet at N=54. Along the star-lattice line the optimized 3-spin basis extends reachable temperatures from T≈0.3 to T≈0.15, and specific-heat data agree with independent NLCE results at several parameter points, resolving a low-T minimum of C/N. The paper also documents an unusual non-monotonic low-temperature sign behavior off the J_t=J_h line and gives a physical explanation via the signed/unsigned ground-state mismatch.

Significance. If the results hold — and the central practical claims are directly measured rather than inferred — this is a useful and broadly applicable contribution. The method is cheap (the cost function requires no QMC), generalizes the stoquasticity-based approach of Ref. [32] from orthogonal to full unitary rotations (thereby recovering the Marshall rotation as a special case), and produces concrete, verifiable gains: a factor-of-four reduction in accessible temperature along the star-lattice line (Fig. 7a), a full computational phase diagram of the MLL (Figs. 1, 6), and specific-heat data at the isotropic and star-lattice points that agree with independent NLCE results and resolve a low-T minimum inaccessible to naive bases (Fig. 9). The recovery of the exact S=0 stoquastic basis for the fully frustrated ladder (Fig. 2a) is a clean validation, and the demonstration that genuinely complex (non-orthogonal) optima exist (§3.2) justifies the U(d) generalization. The multi-start histogram (App. B), the analysis of sign recovery in the dimer phase (§4.3), and the external NLCE benchmarking at five parameter points are welcome, falsifiable checks. The paper is appropriately honest about the near

major comments (3)
  1. [§3.3, Fig. 4, Eq. (12)] §3.3, Fig. 4 (gradient-alignment evidence): the claim that ∇S and ∇⟨s⟩ are aligned — which is what licenses S as a surrogate rather than the procedure being multi-start random search followed by QMC selection of the best basis — rests on finite-difference gradients (Eq. (12), ε=0.075) of a noisy observable at two points of a single run on an N=20 triangular ladder. The comparison in the middle panel is visual ('pointing roughly in the same direction'), with no cosine similarity, no error bars on ∂s/∂θ_i, and no statistics over runs or points. Component-wise sign agreement in 32 dimensions is a weak directional test. Given that this claim appears in the abstract ('These minima coincide with optima of the average sign'), it should be quantified: report the cosine similarity ⟨∇S·∇s⟩/(|∇S||∇s|) with Monte Carlo error bars, at several points along several trajectories.
  2. [§4.2, App. A, Figs. 2/6] §4.2 and App. A: the entire landscape/gradient analysis (Figs. 2, 4, 11) is performed for the U(4)×U(4) Givens parametrization on ladders, but the paper's main application — the MLL computational phase diagram, Figs. 1(b) and 6 — uses U(8)×U(8) with the exponential-map parametrization of Eq. (9), for which App. A itself states that 'the geometric meaning of individual parameter directions remains unclear.' No alignment or trajectory check of any kind is shown for the 3-spin clustering. Since the red regions of Fig. 6 are the headline result, at least one analogue of Fig. 2 (⟨s⟩ vs S along an optimization trajectory) for a representative MLL point in the red region is needed to establish that the surrogate mechanism demonstrated on ladders actually operates in the U(8) case. If it does not, the abstract's mechanistic claim should be scoped to where it is verified.
  3. [Abstract, §2.2, §3.2, Fig. 3] Abstract and §2.2, Eq. (11): the abstract states the minima of S 'coincide with optima of the average sign' unconditionally, while the body is appropriately careful — S=0 is sufficient but not necessary (Eq. (11)), and Fig. 3 shows S is not even an ordinal proxy globally (the best-sign basis has higher S than the rung eigenbasis, as the authors note in §3.2). The abstract should be qualified to match the body's own caveats (e.g., 'local minima of S are observed to coincide with local maxima of the sign for the models studied'), otherwise the headline overstates what is demonstrated. This is a wording change, but it is load-bearing because it defines what readers take the method to guarantee.
minor comments (6)
  1. [§3.1, Fig. 2] §3.1, text below Fig. 2: 'the average sign remains near zero and only starts to increase sharply once the cost function falls below a certain threshold (in Figure 2 at S≈10)' — but the horizontal axis of Fig. 2 spans S≈0–1. Presumably a different threshold or axis scale is meant; please reconcile.
  2. [§4.2] §4.2: 'For every (θ,φ) on the grid, we perform multiple optimizations with 100 random initializations and select the basis that yields the best sign performance.' Selection by measured ⟨s⟩ requires an SSE run per candidate basis at every grid point; please state the number of QMC evaluations per grid point and how temperature was chosen for the selection, since this bears on the advertised cost advantage of the noise-free surrogate.
  3. [§3.2, Fig. 6] §3.2: the claim that no crossing of sign–temperature curves between distinct bases is observed, and that basis ranking is system-size independent, is stated only for the triangular ladder; a sentence clarifying whether this was also checked on the MLL would help, since Fig. 6's basis-color map is temperature-dependent (compare T=0.5 and T=0.2 panels).
  4. [§1, §4.1–4.2, App. C] Typos/formatting: 'infamoussign problem' (§1, missing space); repeated 'Interestingly ,' (space before comma, §1 and §4.1); 'clusters life on a bipartite effective superlattice' → 'lie' (§4.2); 'shows there a inflection point' → 'an inflection point' (§4.2); 'better conditioned' landscape claim in App. C is supported by a single 1D line cut through one minimum — the phrase 'energy landscape' in App. C should read 'cost landscape'.
  5. [Figs. 1, 6] Fig. 1(b) and Fig. 6 duplicate essentially the same information at T=0.5; consider merging or clarifying that Fig. 1(b) is a subset of Fig. 6, and state the grid resolution used for the (θ,φ) maps.
  6. [§2.2, references] Ref. [34] (Murota & Todo, 2025) addresses local basis transformations for sign mitigation and appears closely related; a brief comparative sentence in §2.2 on how the present unitary optimization differs would situate the contribution more precisely.

Circularity Check

0 steps flagged

No significant circularity: S(U) is an independent Hamiltonian-only cost; average sign and NLCE benchmarks are measured separately.

full rationale

The derivation chain does not reduce outputs to inputs by construction. Generalized non-stoquasticity S(U) (Eq. 8) is defined solely from off-diagonal phases of the rotated bond Hamiltonian and is evaluated without QMC. Optimization (Adam+LBFGS over local unitaries) therefore does not fit to the average sign. The average sign 〈s〉 is subsequently measured in independent SSE runs in each candidate basis (computational, cluster eigenbases, and optimized bases), and thermodynamic observables are cross-checked against ED and external NLCE. The claim that local minima of S coincide with maxima of 〈s〉 is an empirical numerical observation (Figs. 2–4, App. C), not a definitional identity; the paper itself states that S=0 is only sufficient, not necessary, and that S is not a global ordinal proxy for sign (Eq. 11 and Fig. 3). Self-citations to prior NLCE/MLL work supply comparison data and context, not the load-bearing justification of the optimization method. Extending Hangleiter et al.’s non-stoquasticity cost from O(d) to U(d) is incremental prior-art use, not a self-citation uniqueness import. No fitted-input-as-prediction, ansatz-smuggling, or renaming circularity is present. Weaknesses in the gradient-alignment evidence (thin, ladder-only) are correctness/robustness issues, not circularity.

Axiom & Free-Parameter Ledger

3 free parameters · 5 axioms · 1 invented entities

The central claim rests on standard SSE-QMC and Heisenberg-model machinery, the definition of a new but elementary cost function built from bond-matrix elements, and the empirical working hypothesis that minimizing that cost improves the average sign. No new physical entities are postulated. Free parameters are ordinary optimizer and sampling hyperparameters, not physics fits.

free parameters (3)
  • Adam iteration count and cosine-annealing schedule = ~4000 Adam steps; relative tol 1e-9
    Chosen by hand (~4000 Adam steps then LBFGS to 1e-9 relative improvement); affects which local minimum is reached but is not fitted to physical observables.
  • Number of Haar-random multi-starts = 100
    Typically 100 random initializations per (θ,φ) point; selection of the best-sign basis among them is a procedural choice that influences the reported computational phase diagram.
  • Finite-difference step ε for gradient comparison = 0.075
    ε=0.075 used in Sec. 3.3; only for diagnostic plots, not for the production optimizations.
axioms (5)
  • domain assumption SSE configuration weights and the reweighting formula 〈O〉 = 〈O s〉_|W| / 〈s〉_|W| correctly estimate thermal observables when the average sign is measurable.
    Standard foundation of signed SSE-QMC (Sec. 2.1); assumed throughout.
  • domain assumption Local unitary rotations on fixed small clusters (2-site or 3-site) are expressive enough to meaningfully reduce the sign problem for the models considered.
    Ansatz restriction stated in Sec. 2.2; larger clusters (e.g. hexagons) are acknowledged as desirable but computationally heavy.
  • ad hoc to paper Generalized non-stoquasticity S(U) = Σ_{i≠j} (1/2)(|[U† h_b U]_{ij}| − Re[U† h_b U]_{ij}) is a useful surrogate whose local minima track average-sign maxima.
    Definition Eq. (8); the tracking property is the paper’s working hypothesis, supported numerically but not proven (Sec. 3.3).
  • domain assumption When the effective superlattice is bipartite, independent unitaries on the two sublattices are allowed (Marshall-type staggering).
    Sec. 2.2; required to recover known stoquastic representations.
  • standard math Standard real analysis / unitary-group parametrizations (Givens products, exponential map of Hermitian generators) cover U(d) for the cluster sizes used.
    Appendix A and Sec. 2.2.
invented entities (1)
  • Generalized non-stoquasticity S(U) for complex bond Hamiltonians independent evidence
    purpose: Provide a cheap, size-independent, noise-free cost function whose minimization is claimed to improve the QMC average sign under local unitary basis changes.
    Direct extension of the real non-stoquasticity of Hangleiter et al. to unitary rotations; vanishes iff the bond Hamiltonian is stoquastic. No claim of new physics—only a computational surrogate.

pith-pipeline@v1.2.0-grok45-kimik3 · 26060 in / 3644 out tokens · 68145 ms · 2026-07-31T08:03:19.224737+00:00 · methodology

0 comments
read the original abstract

The sign problem breaks the polynomial scaling of Quantum Monte Carlo methods. We alleviate it by rotation of the local basis of the Hilbert space, such that the phase of off-diagonal matrix elements of the rotated Hamiltonian is minimized. These minima coincide with optima of the average sign. We benchmark our method for frustrated Heisenberg antiferromagnets in one dimension and on the two-dimensional maple-leaf lattice. Our approach reveals more efficient bases with improved sign in a large part of the parameter space for all models considered, enabling us to reach lower temperatures, on par with state-of-the-art numerical linked cluster expansions which we directly compare to.

Figures

Figures reproduced from arXiv: 2607.24679 by David J. Luitz, Julius S. Herz, Mat\'ias G. Gonzalez, Robin Sch\"afer.

Figure 1
Figure 1. Figure 1: (Left) Maple-leaf lattice with its three symmetry-inequivalent couplings: [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: (Left) average QMC sign 〈s〉 as a function of the generalized non￾stoquasticity S(UA ,UB ) along the optimization trajectory for the fully-frustrated lad￾der (shown in the inset) of N = 24 sites at T = J. (Right) the same for the triangular ladder. 3 Application to ladder systems We now apply our optimization method to two different two-leg ladder systems, the fully frustrated and the triangular ladders sho… view at source ↗
Figure 3
Figure 3. Figure 3: (Left) Average sign against temperature for the triangular ladder system [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: (Left) Sign gradient at the middle and end points of the optimization pro [PITH_FULL_IMAGE:figures/full_fig_p010_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Average QMC sign for the maple-leaf lattice on the [PITH_FULL_IMAGE:figures/full_fig_p012_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Computational phase diagram of the MLL: Average QMC sign of the compu [PITH_FULL_IMAGE:figures/full_fig_p013_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Average QMC sign against temperature for the cluster eigenbasis (com [PITH_FULL_IMAGE:figures/full_fig_p014_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: (Left) QMC sign against temperature on the MLL at [PITH_FULL_IMAGE:figures/full_fig_p015_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Comparison of NLCE and QMC data for the specific heat capacity. The [PITH_FULL_IMAGE:figures/full_fig_p016_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Histogram of the average QMC sign values at temperature [PITH_FULL_IMAGE:figures/full_fig_p020_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Non-stoquasticity S and negative value of the average sign −〈s〉 along a straight line through the parameter space starting from the cost minimum, parametrized by δ ∈ R. The sign is calculated for system sizes N = 20 and N = 24. The simulations are carried out at T = J on a triangular ladder. C Landscape In [PITH_FULL_IMAGE:figures/full_fig_p021_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: Comparison of NLCE (see Ref. [66]) and QMC data for the specific heat capacity at three additional parameter points. The simulation was carried out on an N = 54 lattice in the respective best basis. The previous order in a certain expansion scheme is represented by a dotted line in the respective color [PITH_FULL_IMAGE:figures/full_fig_p022_12.png] view at source ↗
Figure 13
Figure 13. Figure 13: The full N = 54 site maple-leaf cluster used in our simulations, with periodic boundary conditions. 22 [PITH_FULL_IMAGE:figures/full_fig_p022_13.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

69 extracted references · 30 canonical work pages

  1. [1]

    H. T. Diep, Frustrated Spin Systems , World Scientific, doi:10.1142/5697 (2005)

    doi:10.1142/5697 2005

  2. [2]

    Balents, Spin liquids in frustrated magnets , Nature 464(7286), 199 (2010), doi:10.1038/nature08917

    L. Balents, Spin liquids in frustrated magnets , Nature 464(7286), 199 (2010), doi:10.1038/nature08917

    doi:10.1038/nature08917 2010

  3. [3]

    Savary and L

    L. Savary and L. Balents, Quantum spin liquids: a review , Reports on Progress in Physics 80(1), 016502 (2016), doi:10.1088/0034-4885/80/1/016502

    doi:10.1088/0034-4885/80/1/016502 2016

  4. [4]

    Fennell, P

    T. Fennell, P. P. Deen, A. R. Wildes, K. Schmalzl, D. Prabhakaran, A. T. Boothroyd, R. J. Aldus, D. F. McMorrow and S. T. Bramwell, Magnetic coulomb phase in the spin ice ho2ti2o7, Science 326(5951), 415 (2009), doi:10.1126/science.1177582

    doi:10.1126/science.1177582 2009

  5. [5]

    coulomb phase

    C. L. Henley, The “coulomb phase” in frustrated systems, Annual Review of Condensed Matter Physics 1(Volume 1, 2010), 179 (2010), doi:10.1146/annurev-conmatphys-070909-104138

    doi:10.1146/annurev-conmatphys-070909-104138 2010

  6. [6]

    Lozano-Gómez, O

    D. Lozano-Gómez, O. Benton, M. J. P. Gingras and H. Yan, An Atlas of Classical Pyrochlore Spin Liquids , doi:10.48550/arXiv.2411.03547 (2025)

    Pith/arXiv arXiv doi:10.48550/arxiv.2411.03547 2025

  7. [7]

    Hermele, M

    M. Hermele, M. P. A. Fisher and L. Balents, Pyrochlore photons: The U(1) spin liquid in a S= 1 2 three-dimensional frustrated magnet , Phys. Rev. B 69, 064404 (2004), doi:10.1103/PhysRevB.69.064404

    doi:10.1103/physrevb.69.064404 2004

  8. [8]

    B. Gao, F. Desrochers, D. W. Tam, D. M. Kirschbaum, P. Steffens, A. Hiess, D. H. Nguyen, Y. Su, S.-W. Cheong, S. Paschen, Y. B. Kim and P. Dai, Neutron scattering and thermodynamic evidence for emergent photons and fractionalization in a pyrochlore spin ice , Nature Physics 21(8), 1203 (2025), doi:10.1038/s41567-025-02922-9

    doi:10.1038/s41567-025-02922-9 2025

  9. [9]

    Castelnovo, R

    C. Castelnovo, R. Moessner and S. L. Sondhi, Magnetic monopoles in spin ice , Nature 451(7174), 42 (2008), doi:10.1038/nature06433

    doi:10.1038/nature06433 2008

  10. [10]

    D. J. P. Morris, D. A. Tennant, S. A. Grigera, B. Klemke, C. Castelnovo, R. Moessner, C. Czternasty, M. Meissner, K. C. Rule, J.-U. Hoffmann, K. Kiefer, S. Gerischer et al., Dirac Strings and Magnetic Monopoles in the Spin Ice Dy2Ti2O7 , Science 326(5951), 411 (2009), doi:10.1126/science.1178868

    doi:10.1126/science.1178868 2009

  11. [11]

    Broholm, R

    C. Broholm, R. J. Cava, S. A. Kivelson, D. G. Nocera, M. R. Norman and T. Senthil, Quantum spin liquids , Science 367(6475), eaay0668 (2020), doi:10.1126/science.aay0668

    doi:10.1126/science.aay0668 2020

  12. [12]

    S. R. White, Density matrix formulation for quantum renormalization groups, Phys. Rev. Lett. 69, 2863 (1992), doi:10.1103/PhysRevLett.69.2863

    doi:10.1103/physrevlett.69.2863 1992

  13. [13]

    Schollw\"ock, The density-matrix renormalization group, Rev

    U. Schollw\"ock, The density-matrix renormalization group, Rev. Mod. Phys. 77, 259 (2005), doi:10.1103/RevModPhys.77.259

    doi:10.1103/revmodphys.77.259 2005

  14. [14]

    U. Schollwöck, The density-matrix renormalization group in the age of matrix product states , Annals of Physics 326(1), 96 (2011), doi:10.1016/j.aop.2010.09.012, January 2011 Special Issue

    doi:10.1016/j.aop.2010.09.012 2011

  15. [15]

    Hagym\'asi, R

    I. Hagym\'asi, R. Sch\"afer, R. Moessner and D. J. Luitz, Possible inversion symmetry breaking in the s=1/2 pyrochlore heisenberg magnet, Phys. Rev. Lett. 126, 117204 (2021), doi:10.1103/PhysRevLett.126.117204

    work page doi:10.1103/physrevlett.126.117204 2021

  16. [16]

    Becca and S

    F. Becca and S. Sorella, Variational Monte Carlo, p. 101–102, Cambridge University Press, doi:10.1017/9781316417041 (2017)

    doi:10.1017/9781316417041 2017

  17. [17]

    Oitmaa, C

    J. Oitmaa, C. Hamer and W. Zheng, Series Expansion Methods for Strongly Interacting Lattice Models, Cambridge University Press, doi:10.1017/CBO9780511584398 (2006)

    work page doi:10.1017/cbo9780511584398 2006

  18. [18]

    Bernu, L

    B. Bernu, L. Pierre, K. Essafi and L. Messio, Effect of perturbations on the kagome S= 1 2 antiferromagnet at all temperatures , Phys. Rev. B 101, 140403 (2020), doi:10.1103/PhysRevB.101.140403

    doi:10.1103/physrevb.101.140403 2020

  19. [19]

    M. G. Gonzalez, B. Bernu, L. Pierre and L. Messio, Ground-state and thermodynamic properties of the spin- 1 2 Heisenberg model on the anisotropic triangular lattice , SciPost Phys. 12, 112 (2022), doi:10.21468/SciPostPhys.12.3.112

    work page doi:10.21468/scipostphys.12.3.112 2022

  20. [20]

    Niggemann, B

    N. Niggemann, B. Sbierski and J. Reuther, Frustrated quantum spins at finite temperature: Pseudo-Majorana functional renormalization group approach , Phys. Rev. B 103, 104431 (2021), doi:10.1103/PhysRevB.103.104431

    work page doi:10.1103/physrevb.103.104431 2021

  21. [21]

    Niggemann, J

    N. Niggemann, J. Reuther and B. Sbierski, Quantitative functional renormalization for three-dimensional quantum Heisenberg models , SciPost Phys. 12, 156 (2022), doi:10.21468/SciPostPhys.12.5.156

    work page doi:10.21468/scipostphys.12.5.156 2022

  22. [22]

    Schneider, J

    B. Schneider, J. Reuther, M. G. Gonzalez, B. Sbierski and N. Niggemann, Temperature flow in pseudo-Majorana functional renormalization for quantum spins , Phys. Rev. B 109, 195109 (2024), doi:10.1103/PhysRevB.109.195109

    work page doi:10.1103/physrevb.109.195109 2024

  23. [23]

    Bippus, B

    F. Bippus, B. Schneider and B. Sbierski, Pseudo-Majorana functional renormalization for frustrated XXZ spin- 1 2 models with field or magnetization along the spin- Z direction at finite temperature , Phys. Rev. B 111, 054420 (2025), doi:10.1103/PhysRevB.111.054420

    work page doi:10.1103/physrevb.111.054420 2025

  24. [24]

    Rigol, T

    M. Rigol, T. Bryant and R. R. P. Singh, Numerical linked-cluster approach to quantum lattice models, Phys. Rev. Lett. 97, 187202 (2006), doi:10.1103/PhysRevLett.97.187202

    doi:10.1103/physrevlett.97.187202 2006

  25. [25]

    Rigol, T

    M. Rigol, T. Bryant and R. R. P. Singh, Numerical linked-cluster algorithms. i. spin systems on square, triangular, and kagom\'e lattices, Phys. Rev. E 75, 061118 (2007), doi:10.1103/PhysRevE.75.061118

    doi:10.1103/physreve.75.061118 2007

  26. [26]

    Rigol, T

    M. Rigol, T. Bryant and R. R. P. Singh, Numerical linked-cluster algorithms. II . t - J models on the square lattice , Phys. Rev. E 75, 061119 (2007), doi:10.1103/PhysRevE.75.061119

    work page doi:10.1103/physreve.75.061119 2007

  27. [27]

    Sch\"afer, I

    R. Sch\"afer, I. Hagym\'asi, R. Moessner and D. J. Luitz, Pyrochlore S= 1 2 Heisenberg antiferromagnet at finite temperature , Phys. Rev. B 102, 054408 (2020), doi:10.1103/PhysRevB.102.054408

    work page doi:10.1103/physrevb.102.054408 2020

  28. [28]

    Sch\"afer, Magnetic Frustration in Three Dimensions , Dissertation, TU Dresden, Dresden, https://nbn-resolving.org/urn:nbn:de:bsz:14-qucosa2-829375 (2022)

    R. Sch\"afer, Magnetic Frustration in Three Dimensions , Dissertation, TU Dresden, Dresden, https://nbn-resolving.org/urn:nbn:de:bsz:14-qucosa2-829375 (2022)

    2022

  29. [29]

    Troyer and U.-J

    M. Troyer and U.-J. Wiese, Computational Complexity and Fundamental Limitations to Fermionic Quantum Monte Carlo Simulations , Phys. Rev. Lett. 94, 170201 (2005), doi:10.1103/PhysRevLett.94.170201

    doi:10.1103/physrevlett.94.170201 2005

  30. [30]

    Nakamura, Vanishing of the negative-sign problem of quantum Monte Carlo simulations in one-dimensional frustrated spin systems , Phys

    T. Nakamura, Vanishing of the negative-sign problem of quantum Monte Carlo simulations in one-dimensional frustrated spin systems , Phys. Rev. B 57, R3197 (1998), doi:10.1103/PhysRevB.57.R3197

    work page doi:10.1103/physrevb.57.r3197 1998

  31. [31]

    Shinaoka, Y

    H. Shinaoka, Y. Nomura, S. Biermann, M. Troyer and P. Werner, Negative sign problem in continuous-time quantum Monte Carlo: Optimal choice of single-particle basis for impurity problems , Phys. Rev. B 92, 195126 (2015), doi:10.1103/PhysRevB.92.195126

    work page doi:10.1103/physrevb.92.195126 2015

  32. [32]

    Hangleiter, I

    D. Hangleiter, I. Roth, D. Nagaj and J. Eisert, Easing the Monte Carlo sign problem , Science Advances 6(33), eabb8341 (2020), doi:10.1126/sciadv.abb8341

    doi:10.1126/sciadv.abb8341 2020

  33. [33]

    Pan and Z

    G. Pan and Z. Y. Meng, The sign problem in quantum Monte Carlo simulations , In T. Chakraborty, ed., Encyclopedia of Condensed Matter Physics (Second Edition), pp. 879--893. Academic Press, Oxford, second edition edn., ISBN 978-0-323-91408-6, doi:10.1016/B978-0-323-90800-9.00095-0 (2024)

    doi:10.1016/b978-0-323-90800-9.00095-0 2024

  34. [34]

    Murota and S

    K. Murota and S. Todo, Local Basis Transformation to Mitigate Negative Sign Problems , doi:10.48550/arXiv.2501.18069 (2025)

    Pith/arXiv arXiv doi:10.48550/arxiv.2501.18069 2025

  35. [35]

    Wynen, E

    J.-L. Wynen, E. Berkowitz, S. Krieg, T. Luu and J. Ostmeyer, Machine learning to alleviate Hubbard-model sign problems , Phys. Rev. B 103, 125153 (2021), doi:10.1103/PhysRevB.103.125153

    work page doi:10.1103/physrevb.103.125153 2021

  36. [36]

    Rodekamp, E

    M. Rodekamp, E. Berkowitz, C. G \"a ntgen, S. Krieg, T. Luu and J. Ostmeyer, Mitigating the Hubbard sign problem with complex-valued neural networks , Phys. Rev. B 106, 125139 (2022), doi:10.1103/PhysRevB.106.125139

    doi:10.1103/physrevb.106.125139 2022

  37. [37]

    Chandrasekharan and U.-J

    S. Chandrasekharan and U.-J. Wiese, Meron-cluster solution of fermion sign problems, Phys. Rev. Lett. 83, 3116 (1999), doi:10.1103/PhysRevLett.83.3116

    work page doi:10.1103/physrevlett.83.3116 1999

  38. [38]

    Li, Y.-F

    Z.-X. Li, Y.-F. Jiang and H. Yao, Solving the fermion sign problem in quantum monte carlo simulations by majorana representation, Phys. Rev. B 91, 241117(R) (2015), doi:10.1103/PhysRevB.91.241117

    doi:10.1103/physrevb.91.241117 2015

  39. [39]

    Alexandru, G

    A. Alexandru, G. m. c. Ba s s ar, P. F. Bedaque and N. C. Warrington, Complex paths around the sign problem, Rev. Mod. Phys. 94, 015006 (2022), doi:10.1103/RevModPhys.94.015006

    doi:10.1103/revmodphys.94.015006 2022

  40. [40]

    F. Alet, K. Damle and S. Pujari, Sign-Problem-Free Monte Carlo Simulation of Certain Frustrated Quantum Magnets , Phys. Rev. Lett. 117, 197203 (2016), doi:10.1103/PhysRevLett.117.197203

    work page doi:10.1103/physrevlett.117.197203 2016

  41. [41]

    Honecker, S

    A. Honecker, S. Wessel, R. Kerkdyk, T. Pruschke, F. Mila and B. Normand, Thermodynamic properties of highly frustrated quantum spin ladders: Influence of many-particle bound states , Phys. Rev. B 93, 054408 (2016), doi:10.1103/PhysRevB.93.054408

    work page doi:10.1103/physrevb.93.054408 2016

  42. [42]

    Wessel, B

    S. Wessel, B. Normand, F. Mila and A. Honecker, Efficient Quantum Monte Carlo simulations of highly frustrated magnets: the frustrated spin-1/2 ladder , SciPost Phys. 3, 005 (2017), doi:10.21468/SciPostPhys.3.1.005

    work page doi:10.21468/scipostphys.3.1.005 2017

  43. [43]

    D'Emidio, S

    J. D'Emidio, S. Wessel and F. Mila, Reduction of the sign problem near T=0 in quantum Monte Carlo simulations , Phys. Rev. B 102, 064420 (2020), doi:10.1103/PhysRevB.102.064420

    work page doi:10.1103/physrevb.102.064420 2020

  44. [44]

    Honecker, L

    A. Honecker, L. Weber, P. Corboz, F. Mila and S. Wessel, Quantum Monte Carlo simulations of highly frustrated magnets in a cluster basis: The two-dimensional Shastry-Sutherland model , Journal of Physics: Conference Series 2207(1), 012032 (2022), doi:10.1088/1742-6596/2207/1/012032

    work page doi:10.1088/1742-6596/2207/1/012032 2022

  45. [45]

    Weber, A

    L. Weber, A. Honecker, B. Normand, P. Corboz, F. Mila and S. Wessel, Quantum Monte Carlo simulations in the trimer basis: first-order transitions and thermal critical points in frustrated trilayer magnets , SciPost Phys. 12, 054 (2022), doi:10.21468/SciPostPhys.12.2.054

    doi:10.21468/scipostphys.12.2.054 2022

  46. [46]

    Weber, N

    L. Weber, N. Caci and S. Wessel, Cluster quantum Monte Carlo study of two-dimensional weakly coupled frustrated trimer antiferromagnets , Phys. Rev. B 106, 035141 (2022), doi:10.1103/PhysRevB.106.035141

    work page doi:10.1103/physrevb.106.035141 2022

  47. [47]

    Marshall, Antiferromagnetism , Proceedings of the Royal Society of London

    W. Marshall, Antiferromagnetism , Proceedings of the Royal Society of London. A. Mathematical and Physical Sciences 232(1188), 48 (1955), doi:10.1098/rspa.1955.0200

    arXiv 1955

  48. [48]

    P. L. Ebert, Y. Iqbal and A. Wietek, Competing Paramagnetic Phases in the Maple-Leaf Heisenberg Antiferromagnet , doi:10.48550/arXiv.2601.05308 (2026)

    doi:10.48550/arxiv.2601.05308 2026

  49. [49]

    A. W. Sandvik, Stochastic Series Expansion Methods , doi:10.48550/arXiv.1909.10591 (2019)

    Pith/arXiv arXiv doi:10.48550/arxiv.1909.10591 1909

  50. [50]

    F. Alet, S. Wessel and M. Troyer, Generalized directed loop method for quantum Monte Carlo simulations , Phys. Rev. E 71, 036706 (2005), doi:10.1103/PhysRevE.71.036706

    work page doi:10.1103/physreve.71.036706 2005

  51. [51]

    E. Y. Loh, J. E. Gubernatis, R. T. Scalettar, S. R. White, D. J. Scalapino and R. L. Sugar, Sign problem in the numerical simulation of many-electron systems , Phys. Rev. B 41, 9301 (1990), doi:10.1103/PhysRevB.41.9301

    doi:10.1103/physrevb.41.9301 1990

  52. [52]

    Hatano and M

    N. Hatano and M. Suzuki, Representation basis in quantum monte carlo calculations and the negative-sign problem, Physics Letters A 163(4), 246 (1992), doi:10.1016/0375-9601(92)91006-D

    work page doi:10.1016/0375-9601(92)91006-d 1992

  53. [53]

    Henelius and A

    P. Henelius and A. W. Sandvik, Sign problem in Monte Carlo simulations of frustrated quantum spin systems , Phys. Rev. B 62, 1102 (2000), doi:10.1103/PhysRevB.62.1102

    work page doi:10.1103/physrevb.62.1102 2000

  54. [54]

    Levy and B

    R. Levy and B. K. Clark, Mitigating the sign problem through basis rotations, Phys. Rev. Lett. 126, 216401 (2021), doi:10.1103/PhysRevLett.126.216401

    work page doi:10.1103/physrevlett.126.216401 2021

  55. [55]

    Klassen and B

    J. Klassen and B. M. Terhal, Two-local qubit H amiltonians: when are they stoquastic? , Quantum 3, 139 (2019), doi:10.22331/q-2019-05-06-139

    doi:10.22331/q-2019-05-06-139 2019

  56. [56]

    W. R. Clements, P. C. Humphreys, B. J. Metcalf, W. S. Kolthammer and I. A. Walmsley, Optimal design for universal multiport interferometers , Optica 3(12), 1460 (2016), doi:10.1364/OPTICA.3.001460, 1603.08788

    Pith/arXiv arXiv 2016

  57. [57]

    M. Reck, A. Zeilinger, H. J. Bernstein and P. Bertani, Experimental realization of any discrete unitary operator , Phys. Rev. Lett. 73, 58 (1994), doi:10.1103/PhysRevLett.73.58

    doi:10.1103/physrevlett.73.58 1994

  58. [58]

    D. P. Kingma and J. Ba, Adam: A Method for Stochastic Optimization , doi:10.48550/arXiv.1412.6980 (2017)

    Pith/arXiv arXiv doi:10.48550/arxiv.1412.6980 2017

  59. [59]

    D. C. Liu and J. Nocedal, On the limited memory BFGS method for large scale optimization , Mathematical Programming 45(1), 503 (1989), doi:10.1007/BF01589116

    doi:10.1007/bf01589116 1989

  60. [60]

    Mezzadri, How to generate random matrices from the classical compact groups, doi:10.48550/arXiv.math-ph/0609050 (2007)

    F. Mezzadri, How to generate random matrices from the classical compact groups, doi:10.48550/arXiv.math-ph/0609050 (2007)

    Pith/arXiv arXiv doi:10.48550/arxiv.math-ph/0609050 2007

  61. [61]

    Ghosh, T

    P. Ghosh, T. M\"uller and R. Thomale, Another exact ground state of a two-dimensional quantum antiferromagnet , Phys. Rev. B 105, L180412 (2022), doi:10.1103/PhysRevB.105.L180412

    work page doi:10.1103/physrevb.105.l180412 2022

  62. [62]

    Adhikary, A

    M. Adhikary, A. Ralko and B. Kumar, Quantum paramagnetism and magnetization plateaus in a kagome-honeycomb Heisenberg antiferromagnet , Phys. Rev. B 104, 094416 (2021), doi:10.1103/PhysRevB.104.094416

    work page doi:10.1103/physrevb.104.094416 2021

  63. [63]

    Reingruber, N

    A. Reingruber, N. Caci, S. Wessel and J. Richter, Thermodynamics of the spin- 1 2 Heisenberg antiferromagnet on the star lattice , Phys. Rev. B 109, 125120 (2024), doi:10.1103/PhysRevB.109.125120

    work page doi:10.1103/physrevb.109.125120 2024

  64. [64]

    Sonnenschein, A

    J. Sonnenschein, A. Maity, C. Liu, R. Thomale, F. Ferrari and Y. Iqbal, Candidate quantum spin liquids on the maple-leaf lattice, Phys. Rev. B 110, 014414 (2024), doi:10.1103/PhysRevB.110.014414

    work page doi:10.1103/physrevb.110.014414 2024

  65. [65]

    Wessel, I

    S. Wessel, I. Niesen, J. Stapmanns, B. Normand, F. Mila, P. Corboz and A. Honecker, Thermodynamic properties of the shastry-sutherland model from quantum monte carlo simulations, Phys. Rev. B 98, 174432 (2018), doi:10.1103/PhysRevB.98.174432

    work page doi:10.1103/physrevb.98.174432 2018

  66. [66]

    Schäfer, P

    R. Schäfer, P. L. Ebert, N. Hassan, J. Reuther, D. J. Luitz and A. Wietek, Thermodynamics of the heisenberg antiferromagnet on the maple-leaf lattice, Zeitschrift für Naturforschung A (2026), doi:10.1515/zna-2025-0382

    doi:10.1515/zna-2025-0382 2026

  67. [67]

    B. Tang, E. Khatami and M. Rigol, A short introduction to numerical linked-cluster expansions, Computer Physics Communications 184(3), 557 (2013), doi:10.1016/j.cpc.2012.10.008

    work page doi:10.1016/j.cpc.2012.10.008 2013

  68. [68]

    Machado, S

    F. Machado, S. Chern, M. P. Zaletel and N. Y. Yao, A dipolar chiral spin liquid on the breathed kagome lattice, doi:10.48550/arXiv.2603.25784 (2026)

    Pith/arXiv arXiv doi:10.48550/arxiv.2603.25784 2026

  69. [69]

    A. B. Harris, A. J. Berlinsky and C. Bruder, Ordering by quantum fluctuations in a strongly frustrated heisenberg antiferromagnet, Journal of Applied Physics 69(8), 5200 (1991), doi:10.1063/1.348098

    work page doi:10.1063/1.348098 1991