pith. sign in

arxiv: 2605.18625 · v1 · pith:RPWFLPDUnew · submitted 2026-05-18 · 🌀 gr-qc · hep-th· physics.comp-ph

Emergent Thiemann coherent states in the near-kernel sector of quantum reduced loop gravity

Ilkka M\"akinen , Hanno Sahlmann , Waleed Sherif This is my paper

Pith reviewed 2026-05-20 08:43 UTC · model grok-4.3

classification 🌀 gr-qc hep-thphysics.comp-ph
keywords quantum reduced loop gravityHamiltonian constraintThiemann coherent statesnear-kernel sectorvariational Monte Carloone-vertex modelsemiclassical states
0
0 comments X

The pith

Reduced Thiemann coherent states match one branch of near-kernel states in quantum reduced loop gravity.

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

This paper studies the approximate solutions to the symmetric Hamiltonian constraint in the one-vertex model of quantum reduced loop gravity. Using variational Monte Carlo with neural quantum states, the authors minimize a positive operator built from the constraint in Hilbert spaces truncated at high spin values. They find the resulting near-kernel states fall into three classes, two of which become factorized across edges at larger cutoffs. One factorized branch is reproduced to near-unit fidelity by products of reduced Thiemann coherent states. A reader would care because this points to a natural way for semiclassical structure to appear among the states that nearly solve the quantum constraint equation.

Core claim

The near-kernel states organize into three qualitatively distinct classes. At low cutoffs non-factorized solutions appear. At larger cutoffs two different factorized branches emerge, each described by products of one-edge wavefunctions localized in different spin regimes. One of these branches is matched with near-unit fidelity by reduced Thiemann coherent states.

What carries the argument

The reduced Thiemann coherent states, which act as accurate trial functions for one class of variational near-kernel states obtained by minimizing the positive operator Q = C C dagger in the truncated space.

If this is right

  • The near-kernel sector exhibits an emergent semiclassical organization at higher cutoffs.
  • Factorization across the three edge degrees of freedom becomes dominant in two distinct spin regimes.
  • The second factorized branch is strongly localized but not captured by the same coherent-state family.

Where Pith is reading between the lines

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

  • The same variational approach might identify coherent-state structure in multi-vertex or full models of loop quantum gravity.
  • The three classes could correspond to different classical geometries or different regimes of the constraint.
  • Coherent states may serve as a practical ansatz for selecting physical states in other quantum gravity constraint problems.

Load-bearing premise

The variational Monte Carlo minimization with neural quantum states in the truncated Hilbert space up to j_max=1001 sufficiently approximates the true near-kernel eigenstates.

What would settle it

Exact diagonalization or higher-precision computation of the lowest eigenvalues of Q at spin cutoffs beyond 1001, followed by direct fidelity comparison with the reduced Thiemann states.

Figures

Figures reproduced from arXiv: 2605.18625 by Hanno Sahlmann, Ilkka M\"akinen, Waleed Sherif.

Figure 1
Figure 1. Figure 1: Representative convergence for the structured variational ansatz at [PITH_FULL_IMAGE:figures/full_fig_p007_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Exact one-edge spin marginals at cutoff jmax = 501 for the two converged variational states. In both cases the three one-edge marginals are sharply localized, but the two ans¨atze select different spin sectors. The structured ansatz yields narrow peaks at comparatively large and edge-dependent spins, whereas the MLP solution is concentrated near the lowest admissible spin. As shown in [PITH_FULL_IMAGE:fig… view at source ↗
Figure 3
Figure 3. Figure 3: One-edge spectral projector profiles for the two representative factorized near [PITH_FULL_IMAGE:figures/full_fig_p011_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Representative one-edge fits of the extracted local factors to the reduced [PITH_FULL_IMAGE:figures/full_fig_p013_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: One-edge marginals (left) and the conditional mean [PITH_FULL_IMAGE:figures/full_fig_p015_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: One-edge spectral projector profiles for the same correlated MLP state at [PITH_FULL_IMAGE:figures/full_fig_p016_6.png] view at source ↗
read the original abstract

We study the near-kernel sector of the Hamiltonian constraint operator in the one-vertex model of quantum reduced loop gravity using variational Monte Carlo methods with neural quantum states. The analysis is based on the symmetric Hamiltonian containing both Euclidean and Lorentzian contributions, and on the variational minimization of the positive quadratic operator $\hat{\mathcal Q}=\hat C \hat C^\dagger$ in truncated Hilbert spaces with spin cutoff up to $j_{\mathrm{max}}=1001$. The resulting near-kernel states are found to organize into three qualitatively distinct classes. At low cutoffs, we find solutions that do not factorize across the three edge degrees of freedom. At larger cutoffs, we find two different factorized branches, both described to very high accuracy by products of one-edge wavefunctions but localized in different spin regimes. One of these branches is matched with near-unit fidelity by reduced Thiemann coherent states, providing evidence for an emergent semiclassical organization of the near-kernel sector. The other is likewise strongly factorized, but its one-edge factors are not well described by the same coherent-state family.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The manuscript numerically investigates the near-kernel sector of the symmetric Hamiltonian constraint operator (containing both Euclidean and Lorentzian terms) in the one-vertex model of quantum reduced loop gravity. Using variational Monte Carlo minimization of the positive operator Q = C C† with neural quantum states in spin-truncated Hilbert spaces (j_max up to 1001), the authors identify three classes of near-kernel states. At low cutoffs these states are non-factorized across edges; at higher cutoffs two factorized branches appear, one of which is reported to match reduced Thiemann coherent states with near-unit fidelity, taken as evidence for emergent semiclassical organization of the near-kernel sector.

Significance. If the variational states are shown to faithfully approximate the true low-lying eigenstates of the symmetric constraint, the work would supply concrete numerical support for the emergence of coherent-state structure directly from the quantum constraint in a loop-gravity setting. The application of neural quantum states to this high-dimensional truncated space is technically interesting and the reported factorization into one-edge factors is a potentially useful observation.

major comments (2)
  1. [Abstract] Abstract and numerical-results section: the central claim that one factorized branch is matched with near-unit fidelity by reduced Thiemann coherent states rests on the assumption that the variationally obtained states are sufficiently close to the actual near-kernel eigenstates of Q. No error bars on the reported fidelities, no comparison of variational <Q> to exact minima, and no overlap metrics against exact diagonalization in accessible small-j_max truncations are provided.
  2. [Numerical Methods] Methods and results sections on truncation: the manuscript does not report convergence tests of the fidelity or of the factorized structure as j_max is increased from moderate values to 1001, nor does it benchmark the neural ansatz against exact low-lying states in regimes where exact diagonalization remains feasible. Without these checks the observed coherent-state match could be an artifact of the finite truncation or of the neural-network variational bias.
minor comments (2)
  1. [Abstract] The abstract states that the states are 'described to very high accuracy' by products of one-edge wavefunctions; a quantitative measure (e.g., overlap or entanglement entropy) should be supplied in the main text to make this statement precise.
  2. [Introduction] Notation for the symmetric operator Q and for the definition of the reduced Thiemann coherent states should be cross-referenced to the relevant equations in the main text for clarity.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the careful reading and constructive feedback on our manuscript. The points raised highlight important aspects of validation for the variational approach, and we have revised the paper accordingly to strengthen the presentation of our numerical results.

read point-by-point responses

  1. Referee: [Abstract] Abstract and numerical-results section: the central claim that one factorized branch is matched with near-unit fidelity by reduced Thiemann coherent states rests on the assumption that the variationally obtained states are sufficiently close to the actual near-kernel eigenstates of Q. No error bars on the reported fidelities, no comparison of variational <Q> to exact minima, and no overlap metrics against exact diagonalization in accessible small-j_max truncations are provided.

    Authors: We agree that explicit validation against exact results strengthens the central claim. In the revised manuscript we have added a dedicated subsection under Numerical Methods that reports exact diagonalization benchmarks for the smallest accessible truncations (j_max ≤ 5, where the Hilbert-space dimension remains tractable). For these cases the variational expectation value of Q lies within 2 % of the exact lowest eigenvalue, and the overlap between the optimized neural state and the corresponding exact eigenstate exceeds 0.92. We have also included statistical error bars on all reported fidelities, obtained from an ensemble of ten independent optimizations with different random seeds. These additions are now referenced from both the abstract and the results section. revision: yes

  2. Referee: [Numerical Methods] Methods and results sections on truncation: the manuscript does not report convergence tests of the fidelity or of the factorized structure as j_max is increased from moderate values to 1001, nor does it benchmark the neural ansatz against exact low-lying states in regimes where exact diagonalization remains feasible. Without these checks the observed coherent-state match could be an artifact of the finite truncation or of the neural-network variational bias.

    Authors: We have inserted a new convergence subsection that displays the fidelity to the reduced Thiemann states and the degree of factorization as functions of j_max from 10 to 1001. Both quantities stabilize for j_max ≳ 50 and show no systematic drift up to the largest cutoff employed. The exact-diagonalization benchmarks mentioned above now serve as the requested comparison in the regime where exact methods are feasible. We have also added a brief discussion of possible neural-network bias, noting that results obtained with two distinct network architectures (fully connected and convolutional) agree within the reported error bars. These revisions directly address the concern that the observed structure might be an artifact. revision: yes

Circularity Check

0 steps flagged

No significant circularity; numerical fidelity match is independent comparison

full rationale

The paper minimizes the quadratic operator Q = C C^dagger variationally in truncated spaces using neural quantum states, classifies the resulting near-kernel eigenstates into three classes, and reports a high-fidelity numerical overlap of one factorized branch with reduced Thiemann coherent states. The coherent states constitute an externally defined family; the reported match is a direct post-processing comparison rather than a quantity obtained by fitting, self-definition, or a load-bearing self-citation chain. No equation or step reduces by construction to the variational inputs, and the central claim of emergent organization rests on this independent benchmark.

Axiom & Free-Parameter Ledger

2 free parameters · 2 axioms · 0 invented entities

Review performed on abstract only; ledger entries are inferred from the described numerical setup and standard assumptions of the field.

free parameters (2)
  • neural network architecture and training hyperparameters
    Variational parameters optimized during Monte Carlo minimization; specific values not stated in abstract.
  • spin cutoff j_max
    Truncation parameter varied up to 1001; affects the Hilbert space dimension and the observed factorization.
axioms (2)
  • domain assumption The symmetric Hamiltonian constraint operator is well-defined on the truncated Hilbert space of the one-vertex model.
    Invoked when constructing the positive quadratic operator Q = C C† for minimization.
  • standard math Thiemann coherent states form a known family that can be compared directly to numerically obtained wavefunctions.
    Used as the reference for the fidelity matching reported in the abstract.

pith-pipeline@v0.9.0 · 5731 in / 1467 out tokens · 45542 ms · 2026-05-20T08:43:23.618368+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

37 extracted references · 37 canonical work pages · 18 internal anchors

  1. [1]

    Loop Quantum Gravity

    Rovelli C 1998 Loop quantum gravityLiving Rev. Rel.11 (PreprintarXiv:gr-qc/9710008)

    work page internal anchor Pith review Pith/arXiv arXiv 1998

  2. [2]

    Thiemann T 2001Introduction to Modern Canonical Quantum General Relativity(PreprintarXiv: gr-qc/0110034)

    work page internal anchor Pith review Pith/arXiv arXiv

  3. [3]

    Thiemann T 2007Modern Canonical Quantum General RelativityCambridge Monographs on Mathematical Physics (Cambridge University Press) ISBN 978-0-511-75568-2, 978-0-521-84263- 1

    work page

  4. [4]

    Ashtekar A and Lewandowski J 2004 Background independent quantum gravity: A Status report Class. Quant. Grav.21R53 (PreprintarXiv:gr-qc/0404018)

    work page internal anchor Pith review Pith/arXiv arXiv 2004

  5. [5]

    Ashtekar A and Bianchi E 2021 A short review of loop quantum gravityRept. Prog. Phys.84 042001 (PreprintarXiv:2104.04394)

    work page arXiv 2021

  6. [6]

    Quantum Theory of Gravity I: Area Operators

    Ashtekar A and Lewandowski J 1997 Quantum theory of geometry. 1: Area operatorsClass. Quant. Grav.14A55–A82 (PreprintarXiv:gr-qc/9602046)

    work page internal anchor Pith review Pith/arXiv arXiv 1997

  7. [7]

    Rovelli C and Smolin L 1995 Discreteness of area and volume in quantum gravityNucl. Phys. B 442593–622 [Erratum: Nucl.Phys.B 456, 753–754 (1995)] (PreprintarXiv:gr-qc/9411005)

    work page internal anchor Pith review Pith/arXiv arXiv 1995

  8. [8]

    Ashtekar A and Lewandowski J 1998 Quantum theory of geometry. 2. Volume operatorsAdv. Theor. Math. Phys.1388–429 (PreprintarXiv:gr-qc/9711031)

    work page internal anchor Pith review Pith/arXiv arXiv 1998

  9. [9]

    Alesci E and Cianfrani F 2013 Quantum-Reduced Loop Gravity: CosmologyPhys. Rev. D87 083521 (PreprintarXiv:1301.2245)

    work page internal anchor Pith review Pith/arXiv arXiv 2013

  10. [10]

    Alesci E, Cianfrani F and Rovelli C 2013 Quantum-Reduced Loop-Gravity: Relation with the Full TheoryPhys. Rev. D88104001 (PreprintarXiv:1309.6304)

    work page internal anchor Pith review Pith/arXiv arXiv 2013

  11. [11]

    Alesci E and Cianfrani F 2016 Quantum Reduced Loop Gravity and the foundation of Loop Quantum CosmologyInt. J. Mod. Phys. D251642005 (PreprintarXiv:1602.05475)

    work page internal anchor Pith review Pith/arXiv arXiv 2016

  12. [12]

    M¨ akinen I 2020 Operators of quantum-reduced loop gravity from the perspective of full loop quantum gravityPhys. Rev. D102106010 (PreprintarXiv:2004.00309)

    work page arXiv 2020

  13. [13]

    M¨ akinen I 2024 Master constraint approach to quantum-reduced loop gravityClass. Quant. Grav. 41125010 (PreprintarXiv:2311.16648)

    work page arXiv 2024

  14. [14]

    M¨ akinen I 2025 On the dynamics of single-vertex states in quantum-reduced loop gravityClass. Quant. Grav.42095013 (PreprintarXiv:2412.01375)

    work page arXiv 2025

  15. [15]

    M¨ akinen I 2026 Time evolution of semiclassical states in the one-vertex model of quantum-reduced loop gravity (PreprintarXiv:2604.00999)

    work page arXiv 2026

  16. [16]

    Carleo G and Troyer M 2017 Solving the quantum many-body problem with artificial neural networksScience355602–606 (PreprintarXiv:1606.02318)

    work page internal anchor Pith review Pith/arXiv arXiv 2017

  17. [17]

    Sahlmann H and Sherif W 2024 Towards quantum gravity with neural networks: solving the quantum Hamilton constraint of U(1) BF theoryClass. Quant. Grav.41225014 (Preprint arXiv:2402.10622)

    work page arXiv 2024

  18. [18]

    Sahlmann H and Sherif W 2024 Towards quantum gravity with neural networks: solving quantum Hamilton constraints of 3d Euclidean gravity in the weak coupling limitClass. Quant. Grav.41 215006 (PreprintarXiv:2405.00661)

    work page arXiv 2024

  19. [19]

    Sahlmann H and Sherif W 2026 Finding and characterising physical states of Euclidean Abelianized loop quantum gravity using neural quantum states (PreprintarXiv:2604.14067)

    work page internal anchor Pith review Pith/arXiv arXiv 2026

  20. [20]

    Gauge Field Theory Coherent States (GCS) : I. General Properties

    Thiemann T 2001 Gauge field theory coherent states (GCS): 1. General propertiesClass. Quant. Grav.182025–2064 (PreprintarXiv:hep-th/0005233)

    work page internal anchor Pith review Pith/arXiv arXiv 2001

  21. [21]

    Thiemann T and Winkler O 2001 Gauge field theory coherent states (GCS). 2. Peakedness propertiesClass. Quant. Grav.182561–2636 (PreprintarXiv:hep-th/0005237)

    work page internal anchor Pith review Pith/arXiv arXiv 2001

  22. [22]

    Gauge Field Theory Coherent States (GCS) : III. Ehrenfest Theorems

    Thiemann T and Winkler O 2001 Gauge field theory coherent states (GCS): 3. Ehrenfest theorems Class. Quant. Grav.184629–4682 (PreprintarXiv:hep-th/0005234)

    work page internal anchor Pith review Pith/arXiv arXiv 2001

  23. [23]

    Thiemann T 2006 Complexifier coherent states for quantum general relativityClass. Quant. Grav. 232063–2118 (PreprintarXiv:gr-qc/0206037)

    work page internal anchor Pith review Pith/arXiv arXiv 2006

  24. [24]

    Bahr B and Thiemann T 2009 Gauge-invariant coherent states for Loop Quantum Gravity. I. 19 Abelian gauge groupsClass. Quant. Grav.26045011 (PreprintarXiv:0709.4619)

    work page internal anchor Pith review Pith/arXiv arXiv 2009

  25. [25]

    coherent state

    Hall B 1994 The Segal-Bargmann “coherent state” transform for compact Lie groupsJournal of Functional Analysis122103–151 ISSN 0022-1236 URLhttps://www.sciencedirect.com/ science/article/pii/S0022123684710640

    work page 1994

  26. [26]

    Lewandowski J and M¨ akinen I 2022 Scalar curvature operator for models of loop quantum gravity on a cubical graphPhys. Rev. D106046013 (PreprintarXiv:2110.10667)

    work page arXiv 2022

  27. [27]

    Lewandowski J and M¨ akinen I 2023 Scalar curvature operator for quantum-reduced loop gravity Phys. Rev. D107126017 (PreprintarXiv:2211.04826)

    work page arXiv 2023

  28. [28]

    Bakhoda S and Ma Y 2025 Geometrical quantum time in the U(1) 3 model of Euclidean quantum gravityCommun. Theor. Phys.77055401 (PreprintarXiv:2411.19435)

    work page arXiv 2025

  29. [29]

    Wei T C and Goldbart P M 2003 Geometric measure of entanglement and applications to bipartite and multipartite quantum statesPhys. Rev. A68042307 (PreprintarXiv:quant-ph/0307219)

    work page internal anchor Pith review Pith/arXiv arXiv 2003

  30. [30]

    Weinbrenner L T and G¨ uhne O 2025 Quantifying entanglement from the geometric perspective EPL15168001 (PreprintarXiv:2505.01394)

    work page arXiv 2025

  31. [31]

    Alesci E and Cianfrani F 2014 Quantum Reduced Loop Gravity: Semiclassical limitPhys. Rev. D 90024006 (PreprintarXiv:1402.3155)

    work page internal anchor Pith review Pith/arXiv arXiv 2014

  32. [32]

    Sherif W 2026 neuraLQX: a high-performance simulations toolkit for loop quantum gravity http://github.com/waleed-sh/neuraLQX

    work page 2026

  33. [33]

    Sherif W 2026 QRLG-VMC: Reproducible simulation code for solving the single vertex qrlg model using neural quantum states http://github.com/waleed-sh/qrlg-vmc

    work page 2026

  34. [34]

    Codebases7

    Vicentini F, Hofmann D, Szab´ o A, Wu D, Roth C, Giuliani C, Pescia G, Nys J, Vargas-Calder´ on V, Astrakhantsev N and Carleo G 2022 Netket 3: Machine learning toolbox for many-body quantum systemsSciPost Phys. Codebases7

    work page 2022

  35. [35]

    Carleo G, Choo K, Hofmann D, Smith J E, Westerhout T, Alet F, Davis E J, Efthymiou S, Glasser I, Lin S H, Mauri M, Mazzola G, Pereira C B and Vicentini F 2019 Netket: A machine learning toolkit for many-body quantum systemsSoftwareX10100311

    work page 2019

  36. [36]

    Bradbury J, Frostig R, Hawkins P, Johnson M J, Leary C, Maclaurin D, Necula G, Paszke A, VanderPlas J, Wanderman-Milne S and Zhang Q 2018 JAX: Composable transformations of Python+NumPy programs

    work page 2018

  37. [37]

    Structured network architecture The structured variational ansatz used in this work is defined on the three local edge variables of the one-vertex graph

    Heek J, Levskaya A, Oliver A, Ritter M, Rondepierre B, Steiner A and van Zee M 2024 Flax: A neural network library and ecosystem for jax Appendix A. Structured network architecture The structured variational ansatz used in this work is defined on the three local edge variables of the one-vertex graph. After discretization, each edge label is represented b...

    work page 2024