Pith. sign in

REVIEW 4 major objections 3 minor 1 cited by

Taming Landau level mixing in fractional quantum Hall states with deep learning

T0 review · 4 major / 3 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read A real-space neural network wavefunction lowers the variational energy of fractional quantum Hall states below lowest-Landau-level exact diagonalization by naturally including Landau level mixing.

desk verdict Real variational numbers, but the LLM claim outruns the evidence: no multi-LL benchmark to show the energy gain is physical. read the letter →

arxiv 2412.14795 v1 pith:SNUFON4Z submitted 2024-12-19 cond-mat.str-el cond-mat.mes-hallphysics.comp-ph

Yubing Qian , Tongzhou Zhao , Jianxiao Zhang , Tao Xiang , Xiang Li , Ji Chen This is my paper · ORCID
classification cond-mat.str-elcond-mat.mes-hallphysics.comp-ph PACS 73.43.-f71.10.Pm02.70.Ss
keywords fractionalquantumHalleffectLandaulevelmixingneuralnetworkwavefunctionvariationalMonteCarlosphericalgeometryLaughlinstatecompositefermiontransportgap
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper claims that a real-space neural network wavefunction, built from a single determinant of permutation-equivariant multi-electron orbitals multiplied by a Jastrow factor, can represent fractional quantum Hall ground states with Landau level mixing included. For both the ν=1/3 and ν=2/5 fillings on a sphere, the variational energies are consistently lower than exact diagonalization restricted to the lowest Landau level, for systems up to N=12 electrons at ν=1/3 and N=8 at ν=2/5. The energy gap between the neural network result and the lowest-Landau-level exact diagonalization grows monotonically with the Landau level mixing parameter κ, and the network also yields observables such as pair correlation functions, quasiparticle/quasihole densities, and transport gaps that show the expected trends with mixing. A sympathetic reader would care because Landau level mixing is notoriously difficult to treat with traditional methods, and if the neural ansatz truly captures it, the approach opens a practical route to studying fractional quantum Hall physics in the experimentally relevant regime where the Coulomb interaction is comparable to the cyclotron energy.

What carries the argument

The load-bearing object is the single-determinant neural network wavefunction ψ_T = e^J det[φ_i(r_j; {r≠j})], where the Jastrow factor e^J satisfies the electron-electron Coulomb cusp and the multi-electron orbitals φ_i(r_j; {r≠j}) = Σ_{k,m} w_{ikm} f_k(r_j; {r≠j}) $u_j^{{Q+m}}$ $v_j^{{Q-m}}$ combine LLL monopole harmonics (defined through spinor coordinates on the sphere, which handle the Dirac-string phase) with permutation-equivariant features f_k produced by the Psiformer, a self-attention transformer architecture. The neural network features encode many-body correlations, and because they are not projected onto the LLL, the product with monopole harmonics yields components in higher Landau levels, giving the ansatz a natural mechanism for Landau level mixing. The determinant enforces antisymmetry while the Jastrow factor controls short-range behavior; variational Monte Carlo with the Kronecker-Factored Approximate Curvature optimizer minimizes the energy. This construction avoids the local-energy divergence at the poles that would arise from a bare neural-network wavefunction on the sphere.

What would settle it

For a small system (e.g., N=4 or N=6 electrons at ν=1/3 on the sphere, same 2Q as in the paper), perform exact diagonalization in a Hilbert space that explicitly includes the lowest two or three Landau levels at κ=1 and compare the ground-state energy and the lowest-Landau-level occupation with the neural network result. If the neural network energy is not equal to or lower than the multi-Landau-level ED energy within statistical error, or if its inferred N_LLL disagrees with the ED value, the claim that the ansatz naturally captures Landau level mixing would be contradicted.

Watch

Extended reading notes

Core claim

The central discovery is that the neural network variational ansatz of Eqs. (3)-(5) can go beyond the lowest Landau level approximation without explicit multi-level construction. The wavefunction ψ_T = e^J det[φ_i(r_j; {r≠j})] uses LLL monopole harmonics multiplied by neural-network features, so the orbitals are not restricted to the LLL and the determinant-plus-Jastrow form can represent higher Landau level admixtures. On the sphere, for ν=1/3 with N=6, 8, 10, 12 electrons and for ν=2/5 with N=8 at κ=1, the optimized energies fall below exact diagonalization results that only include the LLL, and the energy difference increases with κ. The fraction of electrons in the LLL, N_LLL/N, decreases from about 0.999 at κ=0.5 to about 0.96 at κ=10, while the overlap with the Laughlin wavefunction drops from about 0.998 to 0.873, showing that the network is capturing the progressive Landau level mixing. The same ansatz produces quasiparticle and quasihole charge densities, pair correlation functions, and a transport gap that shrinks with κ but stays open up to κ=10, consistent with a gapped FQH phase at moderate mixing.

Load-bearing premise

The variational manifold spanned by the neural network wavefunction is expressive enough that the energy lowering below the lowest-Landau-level exact diagonalization comes from physical Landau level mixing, rather than from generic variational freedom or a systematic bias in the ansatz.

Editorial extensions

If this is right

  • The neural network method can reach system sizes (e.g., N=12 at ν=1/3) that are inaccessible to LLL exact diagonalization, while still including Landau level mixing, so it can extend FQH phase-diagram studies to larger systems and stronger mixing.
  • Because the energy gap between the neural network result and LLL ED grows with κ, the method provides a quantitative handle on how much Landau level mixing affects ground-state energies and observables in the experimentally relevant regime.
  • The computed transport gap decreases with κ but remains open up to κ=10, suggesting that the ν=1/3 FQH state survives moderate Landau level mixing, in agreement with earlier studies that the paper cites.
  • The same ansatz works for the more complex ν=2/5 composite-fermion filling, indicating that the approach may generalize to other Jain states and potentially to non-Abelian FQH states with a modified pairing factor.
  • The pair correlation function develops a peak at θ=π at large κ, hinting at a tendency toward Wigner crystallization, though the paper explicitly notes the system size is too small to pin down the phase boundary.

Reading between the lines

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

  • The central claim could be tested against a multi-Landau-level exact diagonalization or a fixed-phase diffusion Monte Carlo calculation for the same small systems; if the neural network energy matches or lies below such a reference, the physical origin of the energy drop would be confirmed, whereas a systematic disagreement would indicate variational bias rather than true Landau level mixing.
  • Since the ansatz uses a single Slater determinant (rather than a sum of determinants), the expressivity is carried almost entirely by the neural-network orbitals; one could probe this by checking whether the energy converges monotonically as network width and depth grow, and whether the N_LLL value from the network matches perturbative estimates at small κ.
  • The observed N_LLL ≈ 0.96 even at κ=10 implies that only a few percent of electrons leave the LLL, yet this small fraction shifts energies and observables noticeably; a natural extension is to map N_LLL as a function of κ for larger N to see whether the higher-LL occupation saturates or grows further.
  • The methodology may transfer to other geometries such as the disk or torus, where the monopole-harmonic trick would need to be replaced by plane-wave or other bases; the concurrent disk-geometry study cited by the paper suggests the approach is geometry-flexible.
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

4 major / 3 minor

Summary. The manuscript proposes a real-space neural-network variational Monte Carlo (NNVMC) approach for fractional quantum Hall systems on the sphere. The trial wavefunction (Eq. 3) is a Jastrow factor times a single Slater determinant whose multi-electron orbitals are built from LLL monopole harmonics multiplied by permutation-equivariant features from a Psiformer-type network. Parameters are optimized by minimizing the variational energy of the Hamiltonian in Eq. (1), so no exact-diagonalization data are used as labels. For nu=1/3 and 2/5 at kappa=1 with N=6 to 12, the neural-network energies are reported to be consistently lower than LLL-only exact diagonalization (ED). The paper also reports overlap with the Laughlin wavefunction, the number of electrons on the LLL, pair correlation functions, quasiparticle/quasihole densities, and a transport gap as functions of kappa. The central claim is that the real-space network naturally captures Landau level mixing and yields energies below LLL ED.

Significance. If the Landau-level-mixing content were validated, this would be a significant methodological advance: the method provides variational upper bounds below LLL ED, reaches N=12 where LLL ED is expensive, and produces a kappa-dependent transport gap with a plausible Wigner-crystal tendency at large kappa. The paper is honest in presenting the energy as a variational upper bound, and the overlap, NLLL, and pair-correlation diagnostics are evaluated after optimization rather than fit to ED, so there is no obvious circularity. However, the central LLM claim is currently unvalidated: there is no multi-Landau-level benchmark, no reported angular-momentum content of the ground state, and no release of code or data. The significance of the work therefore rests on a load-bearing assumption that the variational manifold selects the physical higher-Landau-level admixture.

major comments (4)
  1. [Abstract; Eqs. (3)-(5); Fig. 3b; Table S3] The central claim that the real-space neural-network wavefunction 'naturally captures the extent of Landau level mixing' is not supported by any multi-Landau-level reference calculation. Energies below LLL ED prove only that the optimized state is not entirely in the LLL; they do not prove that the higher-LL component is the physical one. The only quantitative LLM diagnostic, NLLL in Fig. 3b (and Table S3), is a scalar and cannot distinguish a correct LLM-corrected liquid from a state with spurious higher-LL admixture. I request a systematic comparison with a multi-LL exact diagonalization for N=6 at the same flux (or a fixed-phase DMC / multi-LL DMRG reference), and a decomposition of the one-body density matrix into individual Landau-level occupations, not just their sum.
  2. [Eqs. (3)-(5); Figs. 2-3] The trial state is not constrained to the L=0 sector, although the Hamiltonian (1) is rotationally invariant and the FQH ground states studied are angular-momentum singlets. A finite-size variational ansatz can lower energy by spontaneously breaking rotational symmetry, and the scalar NLLL and overlap diagnostics do not rule this out. The paper never reports the expectation value of L^2 for the ground-state calculations. I ask that the expectation value of L^2 (or an angular-momentum decomposition) be reported for the states underlying Figs. 2-4, and ideally that the ground-state optimization be performed with a projection to L=0, so that the energy gain over LLL ED can be attributed to physical Landau level mixing rather than symmetry breaking.
  3. [Fig. 4c; Table S3; Eq. (A.16)] The transport-gap prediction in Fig. 4c and the conclusion that the gap persists at kappa=10 rest on quasiparticle/quasihole energies that are not validated. Excitation states are selected only by adding a soft Lz penalty with strength beta (Table S1), and no convergence with respect to beta or any check of the L^2 content of the excited states is provided. At small kappa the neural-network gap happens to lie close to the LLL ED value, but LLL ED is not a multi-LL reference, so this does not validate the kappa dependence. Please benchmark the excitation energies against a multi-LL reference at N=6 (or against leading-order LLM perturbation theory), and show that the reported gap is stable with respect to beta and to the choice of target Lz.
  4. [Appendix: Computational details; Fig. S1] The handling of local-energy outliers is described only qualitatively. The manuscript states that outliers are rare and do not influence training, but no specification is given of how the final energy estimates in Tables S2-S3 are computed, for example whether any clipping or rejection is applied to the local-energy samples used for the reported means and error bars. Since the appendix itself explains that the local energy near the poles can diverge for a generic neural-network phase, this should be quantified: please report the histogram of local energies and the sensitivity of the final energies to outlier removal.
minor comments (3)
  1. [Fig. 3a vs Table S3] The unit notation is inconsistent: Fig. 3a uses 'E_c/N hbar omega_c kappa' while Table S3 uses 'E_c/N kappa hbar omega_c'; please unify the notation throughout.
  2. [Eq. (A.19)-(A.20)] The definition of the one-body reduced density matrix in Eq. (A.19) should state explicitly that the monopole harmonics are normalized and that the spinor measure on the sphere is included in the integrals, so that the trace of the 1-RDM equals N.
  3. [Reproducibility] No code or data are released, which makes it difficult to reproduce the variational energies, overlaps, and gap values reported in Tables S2-S3 and Figs. 2-4; providing the optimized wavefunctions or the training code would strengthen the manuscript.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the NNVMC energies are variational upper bounds minimized against the Hamiltonian, the LLL-ED numbers are external benchmarks rather than fitted inputs, and the single same-group citation is only a non-load-bearing consistency check.

full rationale

The paper's derivation chain is variational, not inferential-from-fit. The trial state (Eqs. 3-5) is a single determinant of permutation-equivariant neural-network orbitals built on LLL monopole harmonics times a Jastrow factor; the parameters are optimized by minimizing the Monte Carlo estimate of the full Hamiltonian (Eq. 1), which includes the kinetic Landau level term and Coulomb interaction. The LLL ED benchmark is an external comparison (DiagHam), not an input: no ED eigenvalue, overlap, or gap enters the loss function or defines the ansatz. The reported overlap, NLLL, PCF, and transport gap are evaluated after optimization from the optimized wavefunction, so none is a fitted quantity renamed as a prediction. The transport gap is computed from separately optimized states at fluxes 2Q=14,15,16 with an Lz-pinning penalty, which selects a symmetry sector rather than fitting the gap. The only same-group citation is Ref. [50] in the sentence 'These results are all consistent with previous studies [50,57,58]'; this is a post-hoc consistency check, not a load-bearing premise or a parameter source. No equation in the paper reduces to its own input by construction. The absence of a converged multi-Landau-level DMRG or fixed-phase DMC benchmark is a correctness and validation concern, not a circularity concern.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The central numerical claim uses only standard Haldane-sphere physics; the only hand-fitted quantities are the Jastrow length scale and the excitation-selection penalty. No new particles or forces are introduced. The main unproven input is the expressivity of the NN ansatz, and the paper does not benchmark against a multi-LL reference calculation.

free parameters (2)
  • Jastrow parameter a = not reported
    Controls the Coulomb cusp in Eq. (4); optimized during VMC but its converged value is absent from the paper, making exact reproduction impossible.
  • Angular momentum penalty beta and target Lz = beta=0.02 for quasiparticle, beta=0.01 for quasihole (Table S1); target m_t chosen per sector
    Added to the loss as beta(Lz-mt)^2 to select quasiparticle and quasihole sectors; these hyperparameters are hand-chosen and affect the reported excitation densities.
assumptions (4)
  • domain assumption The Haldane-sphere Hamiltonian (Eq. 1) with spin-polarized electrons, a single band mass m, and Coulomb interaction is the correct model for the FQH systems studied.
    This is the standard model used throughout the FQH literature and defines the target ground states; the paper does not derive it.
  • ad hoc to paper The single-determinant ansatz built from LLL monopole harmonics, permutation-equivariant neural-network features, and a Jastrow factor can represent the exact LLM ground state well enough that the energy gap to LLL ED is physical.
    Eqs. (3)-(5) define the variational manifold. The paper does not prove expressiveness or benchmark against a multi-LL exact or DMRG calculation, so this is the load-bearing unproven input.
  • domain assumption Higher Landau level components are generated implicitly by the non-holomorphic Jastrow factor and the neural-network features rather than by explicit higher-LL orbitals.
    Only LLL monopole harmonics appear in Eq. (5); the claimed ability to naturally include LLM relies on this assumption, and only average NLLL is reported, not a full LL occupation decomposition.
  • domain assumption The finite-size energy corrections in Eqs. (A.13)-(A.16), including background charge, kinetic shift, and density correction, are appropriate for comparing spherical FQH energies.
    These are standard corrections from the composite-fermion literature and are applied identically to NN and ED results, so any residual error affects both sides equally.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Taming Landau level mixing in fractional quantum Hall states with deep learning." pith.science (2026). https://pith.science/paper/SNUFON4Z

@misc{pith2026241214795,
  author       = {Pith},
  title        = {Pith review of: Taming Landau level mixing in fractional quantum Hall states with deep learning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/SNUFON4Z}},
  note         = {Machine review of arXiv:2412.14795}
}
abstract

Strong correlation brings a rich array of emergent phenomena, as well as a daunting challenge to theoretical physics study. In condensed matter physics, the fractional quantum Hall effect is a prominent example of strong correlation, with Landau level mixing being one of the most challenging aspects to address using traditional computational methods. Deep learning real-space neural network wavefunction methods have emerged as promising architectures to describe electron correlations in molecules and materials, but their power has not been fully tested for exotic quantum states. In this work, we employ real-space neural network wavefunction techniques to investigate fractional quantum Hall systems. On both $1/3$ and $2/5$ filling systems, we achieve energies consistently lower than exact diagonalization results which only consider the lowest Landau level. We also demonstrate that the real-space neural network wavefunction can naturally capture the extent of Landau level mixing up to a very high level, overcoming the limitations of traditional methods. Our work underscores the potential of neural networks for future studies of strongly correlated systems and opens new avenues for exploring the rich physics of the fractional quantum Hall effect.

Figures

Figures reproduced from arXiv: 2412.14795 by the authors.

Figure 1
Figure 1. FIG. 1 [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2 [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. FIG. 3 [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: FIG. 4 [PITH_FULL_IMAGE:figures/full_fig_p004_4.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

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

  1. Is attention all you need to solve the correlated electron problem?

    cond-mat.str-el 2025-02 conditional novelty 6.0 of 10

    A self-attention neural network wavefunction gives lower variational energies than band-projected exact diagonalization for a moiré electron model and shows a roughly quadratic parameter scaling with electron number.

Reference graph

Works this paper leans on

62 extracted references · 51 canonical work pages · cited by 1 Pith paper

  1. [1]

    H. L. Stormer, D. C. Tsui, and A. C. Gossard, Reviews of Modern Physics 71, S298 (1999)

    work page 1999

  2. [2]

    R. B. Laughlin, Phys. Rev. Lett. 50, 1395 (1983)

    1983

  3. [3]

    Sodemann and A

    I. Sodemann and A. H. MacDonald, Physical Review B 87, 245425 (2013)

    work page 2013

  4. [4]

    V. J. Goldman, M. Santos, M. Shayegan, and J. E. Cun- ningham, Phys. Rev. Lett. 65, 2189 (1990)

    work page 1990

  5. [5]

    Thiebaut, N

    N. Thiebaut, N. Regnault, and M. O. Goerbig, Phys. Rev. B 92, 245401 (2015)

    work page 2015

  6. [6]

    J. P. Eisenstein, H. L. Stormer, L. Pfeiffer, and K. W. West, Phys. Rev. Lett. 62, 1540 (1989)

    work page 1989

  7. [7]

    D. R. Luhman, W. Pan, D. C. Tsui, L. N. Pfeiffer, K. W. Baldwin, and K. W. West, Phys. Rev. Lett. 101, 266804 (2008)

    work page 2008

  8. [8]

    Y.-L. Wu, B. Estienne, N. Regnault, and B. A. Bernevig, Phys. Rev. Lett. 113, 116801 (2014)

    work page 2014

Show all 62 references
  1. [9]

    Regnault, J

    N. Regnault, J. Maciejko, S. A. Kivelson, and S. L. Sondhi, Phys. Rev. B 96, 035150 (2017) . 6

    2017

  2. [10]

    Yoshioka, Journal of the Physical Society of Japan 53, 3740 (1984)

    D. Yoshioka, Journal of the Physical Society of Japan 53, 3740 (1984)

    1984

  3. [11]

    A. E. Feiguin, E. Rezayi, C. Nayak, and S. Das Sarma, Physical Review Letters 100, 166803 (2008)

    2008

  4. [12]

    M. P. Zaletel, R. S. K. Mong, F. Pollmann, and E. H. Rezayi, Physical Review B 91, 045115 (2015)

    2015

  5. [13]

    Ortiz, D

    G. Ortiz, D. M. Ceperley, and R. M. Martin, Physical Review Letters 71, 2777 (1993)

    1993

  6. [14]

    J. Zhao, Y. Zhang, and J. K. Jain, Phys. Rev. Lett. 121, 116802 (2018)

    2018

  7. [15]

    T. Zhao, A. C. Balram, and J. K. Jain, Phys. Rev. Lett. 130, 186302 (2023)

    2023

  8. [16]

    Hermann, J

    J. Hermann, J. Spencer, K. Choo, A. Mezzacapo, W. M. C. Foulkes, D. Pfau, G. Carleo, and F. No´ e, Na- ture Reviews Chemistry 7, 692 (2023)

    2023

  9. [17]

    Y. Qian, X. Li, Z. Li, W. Ren, and J. Chen, Deep learning quantum Monte Carlo for solids (2024), arXiv:2407.00707 [cond-mat, physics:physics]

    2024 arXiv

  10. [18]

    Carleo and M

    G. Carleo and M. Troyer, Science 355, 602 (2017)

    2017

  11. [19]

    Vicentini, A

    F. Vicentini, A. Biella, N. Regnault, and C. Ciuti, Phys. Rev. Lett. 122, 250503 (2019)

    2019

  12. [20]

    J. Han, L. Zhang, and W. E, Journal of Computational Physics 399, 108929 (2019)

    2019

  13. [21]

    D. Pfau, J. S. Spencer, A. G. D. G. Matthews, and W. M. C. Foulkes, Physical Review Research 2, 033429 (2020)

    2020

  14. [22]

    Hermann, Z

    J. Hermann, Z. Sch¨ atzle, and F. No´ e,Nature Chemistry 12, 891 (2020)

    2020

  15. [23]

    K. Choo, A. Mezzacapo, and G. Carleo, Nature Commu- nications 11, 2368 (2020)

    2020

  16. [24]

    von Glehn, J

    I. von Glehn, J. S. Spencer, and D. Pfau, in The Eleventh International Conference on Learning Representations, ICLR 2023 (OpenReview.net, kigali, rwanda, 2023)

    2023

  17. [25]

    Yoshioka, W

    N. Yoshioka, W. Mizukami, and F. Nori, Communica- tions Physics 4, 1 (2021)

    2021

  18. [26]

    X. Li, Y. Qian, and J. Chen, Physical Review Letters 132, 176401 (2024)

    2024

  19. [27]

    X. Li, Z. Li, and J. Chen, Nature Communications 13, 7895 (2022)

    2022

  20. [28]

    Wilson, S

    M. Wilson, S. Moroni, M. Holzmann, N. Gao, F. Wu- darski, T. Vegge, and A. Bhowmik, Physical Review B 107, 235139 (2023)

    2023

  21. [29]

    Cassella, H

    G. Cassella, H. Sutterud, S. Azadi, N. D. Drummond, D. Pfau, J. S. Spencer, and W. M. C. Foulkes, Physical Review Letters 130, 036401 (2023)

    2023

  22. [30]

    J. Kim, G. Pescia, B. Fore, J. Nys, G. Carleo, S. Gan- dolfi, M. Hjorth-Jensen, and A. Lovato, Communications Physics 7, 148 (2024)

    2024

  23. [31]

    X. Li, Y. Qian, W. Ren, Y. Xu, and J. Chen, Emergent Wigner phases in moir´ e superlattice from deep learning (2024), arXiv:2406.11134 [cond-mat, physics:physics]

    2024 arXiv

  24. [32]

    D. Luo, D. D. Dai, and L. Fu, Simulating moir´ e quantum matter with neural network (2024), arXiv:2406.17645 [cond-mat]

    2024 arXiv

  25. [33]

    F. D. M. Haldane, Phys. Rev. Lett. 51, 605 (1983)

    1983

  26. [34]

    Yoshioka, B

    D. Yoshioka, B. I. Halperin, and P. A. Lee, Phys. Rev. Lett. 50, 1219 (1983)

    1983

  27. [35]

    X. G. Wen and A. Zee, Physical Review Letters 69, 953 (1992)

    1992

  28. [36]

    J. K. Jain, Composite Fermions, 1st ed. (Cambridge Uni- versity Press, Cambridge ; New York, 2007)

    2007

  29. [37]

    R. Morf, N. d’Ambrumenil, and B. I. Halperin, Phys. Rev. B 34, 3037 (1986)

    1986

  30. [38]

    Kato, Communications on Pure and Applied Mathe- matics 10, 151 (1957)

    T. Kato, Communications on Pure and Applied Mathe- matics 10, 151 (1957)

    1957

  31. [39]

    T. T. Wu and C. N. Yang, Nucl. Phys. B 107, 365 (1976)

    1976

  32. [40]

    Martens and R

    J. Martens and R. Grosse, in Proceedings of the 32nd International Conference on Machine Learning (PMLR,

  33. [41]

    J. K. Jain, Phys. Rev. Lett. 63, 199 (1989)

    1989

  34. [42]

    R. K. Kamilla, J. K. Jain, and S. M. Girvin, Physical Review B 56, 12411 (1997)

    1997

  35. [43]

    Yannouleas and U

    C. Yannouleas and U. Landman, Phys. Rev. B 68, 035326 (2003)

    2003

  36. [44]

    Yannouleas and U

    C. Yannouleas and U. Landman, Phys. Rev. B 66, 115315 (2002)

    2002

  37. [45]

    Shibata and D

    N. Shibata and D. Yoshioka, Phys. Rev. Lett. 86, 5755 (2001)

    2001

  38. [46]

    J. J. Thomson, Phil. Mag. 7, 237 (1904)

    1904

  39. [47]

    Wigner, Phys

    E. Wigner, Phys. Rev. 46, 1002 (1934)

    1934

  40. [48]

    F. D. M. Haldane and E. H. Rezayi, Phys. Rev. Lett. 54, 237 (1985)

    1985

  41. [49]

    R. H. Morf, N. d’Ambrumenil, and S. Das Sarma, Phys. Rev. B 66, 075408 (2002)

    2002

  42. [50]

    T. Zhao, K. Kudo, W. N. Faugno, A. C. Balram, and J. K. Jain, Phys. Rev. B 105, 205147 (2022)

    2022

  43. [51]

    G. S. Boebinger, A. M. Chang, H. L. Stormer, and D. C. Tsui, Phys. Rev. Lett. 55, 1606 (1985)

    1985

  44. [52]

    R. R. Du, H. L. Stormer, D. C. Tsui, L. N. Pfeiffer, and K. W. West, Phys. Rev. Lett. 70, 2944 (1993)

    1993

  45. [53]

    B. I. Halperin, P. A. Lee, and N. Read, Phys. Rev. B 47, 7312 (1993)

    1993

  46. [54]

    Murthy and R

    G. Murthy and R. Shankar, Reviews of Modern Physics 75, 1101 (2003)

    2003

  47. [55]

    Yu, Z.-B

    Y. Yu, Z.-B. Su, and X. Dai, Phys. Rev. B 57, 9897 (1998)

    1998

  48. [56]

    Praz, Phys

    A. Praz, Phys. Rev. B 75, 205342 (2007)

    2007

  49. [57]

    Melik-Alaverdian and N

    V. Melik-Alaverdian and N. E. Bonesteel, Phys. Rev. B 52, R17032 (1995)

    1995

  50. [58]

    Murthy and R

    G. Murthy and R. Shankar, Physical Review B 65, 245309 (2002)

    2002

  51. [59]

    Y. Teng, D. D. Dai, and L. Fu, Solving and visualizing fractional quantum Hall wavefunctions with neural net- work (2024), arXiv:2412.00618 [cond-mat]

    2024 arXiv

  52. [60]

    DiagHam, https://nick-ux.org/diagham

  53. [61]

    Fishman, S

    M. Fishman, S. R. White, and E. M. Stoudenmire, Sci- Post Phys. Codebases , 4 (2022) . 7 Appendix: methodological and computational details The role of monopole harmonics It is mentioned in the main text that LLL monopole harmonics uQ+m j vQ−m j are used to avoid the divergenc...

    2022 arXiv

  54. [62]

    9976(2) 5

    5 0 . 9976(2) 5 . 9989(8) −0. 41106(2) −0. 397425(8) −0. 405843(7) 0 . 1130(3) 1 0 . 994(2) 5 . 9955(8) −0. 411645(3) −0. 39897(1) −0. 406667(3) 0 . 1058(2) 3 0 . 975(2) 5 . 9644(8) −0. 413331(3) −0. 403202(1) −0. 408784(1) 0 . 08805(5) 10 0 . 873(1) 5 . 7500(8) −0. 417425(1) ...

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.