pith. sign in

arxiv: 2606.05719 · v1 · pith:XYYE5CUUnew · submitted 2026-06-04 · 🪐 quant-ph · hep-lat

Symmetries and overparametrization properties of Hamiltonian variational ansatzes for the (1+1)d mathbb{Z}₂ lattice gauge theory

Kanta Yamanaka , Takanori Daiza , Katsumi Imaizumi , Yutaro Iiyama , Lento Nagano , Ryu Sawada , Koji Terashi This is my paper

Pith reviewed 2026-06-28 01:32 UTC · model grok-4.3

classification 🪐 quant-ph hep-lat
keywords Hamiltonian variational ansatzoverparametrizationZ2 lattice gauge theoryvariational quantum eigensolverdynamical Lie algebraquantum Fisher information matrixsymmetrieslattice gauge theory
0
0 comments X

The pith

Symmetry-preserving Hamiltonian variational ansatzes for the (1+1)d Z2 lattice gauge theory show overparametrization that removes local minima from the VQE loss and makes loss decay scale linearly with parameter count.

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

The paper examines five Hamiltonian variational ansatzes built from the (1+1)d Z2 lattice gauge theory Hamiltonian and required to respect its local and global symmetries. These ansatzes act inside finely segmented invariant subspaces of the Hilbert space. The authors compute the dimension of the dynamical Lie algebra generated by each ansatz and the rank of its quantum Fisher information matrix. Numerical VQE experiments then reveal that once the parameter count reaches the point of overparametrization, local minima disappear from the loss surface. At the same time the rate at which gradient descent reduces the loss becomes directly proportional to the number of variational parameters.

Core claim

For the five studied Hamiltonian variational ansatzes based on the (1+1)d Z2 lattice gauge theory Hamiltonian, overparametrization coincides with the apparent disappearance of local minima in the VQE loss function, and the decay rate of the loss under gradient descent optimization scales linearly with the number of parameters in the ansatz.

What carries the argument

Hamiltonian variational ansatzes (HVA) that respect local and global symmetries of the Z2 lattice gauge theory, with generators formed from sums of weight-three Pauli operators, whose effective dimension is measured by dynamical Lie algebra dimension and quantum Fisher information matrix rank inside invariant subspaces.

If this is right

  • Overparametrized symmetry-respecting ansatzes avoid local minima during VQE optimization for this lattice gauge model.
  • Gradient-descent convergence speed increases linearly with parameter count once overparametrization is reached.
  • Symmetry enforcement in the ansatz design directly influences the structure of the loss landscape in variational quantum algorithms.
  • The results supply concrete guidance for constructing scalable variational circuits for lattice gauge theories.

Where Pith is reading between the lines

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

  • The linear scaling may occur because overparametrization lets the circuit reach every direction inside the relevant invariant subspace without redundant parameters.
  • Analogous overparametrization thresholds could appear in higher-dimensional or non-Abelian gauge theories provided the generators keep a comparable weight-three Pauli structure.
  • The underexplored use of weight-three Pauli sums may be responsible for the clean transition between under- and over-parametrized regimes.

Load-bearing premise

The five chosen ansatzes and the specific (1+1)d Z2 model are representative enough that the observed overparametrization behavior and linear scaling will generalize to other gauge theories or ansatz families.

What would settle it

A numerical VQE run on a larger lattice or different gauge group in which an ansatz is overparametrized (DLA dimension or QFIM rank saturated) yet local minima remain visible in the loss landscape would falsify the reported coincidence.

Figures

Figures reproduced from arXiv: 2606.05719 by Kanta Yamanaka, Katsumi Imaizumi, Koji Terashi, Lento Nagano, Ryu Sawada, Takanori Daiza, Yutaro Iiyama.

Figure 2
Figure 2. Figure 2: FIG. 2. Algebraic dimensions of the DLA subrepresenta [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: FIG. 3. Maximum QFIM rank [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: FIG. 4. Average decay rate [PITH_FULL_IMAGE:figures/full_fig_p008_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: shows the second frame potential of Vj at ten values of j for the five ansatz families with Ns = 12. The number of ansatz layers is 20 for families A to D, and 30 for family E. The horizontal axis corresponds to the parameter index j, and is plotted in reverse order to show the results for deeper layers on the right (see the definition of V+j in Eq. (B7)). In families A to C, F2 indeed becomes consistent w… view at source ↗
read the original abstract

We perform detailed studies of five Hamiltonian variational ansatzes (HVA) based on the Hamiltonian of the $(1+1)$d $\mathbb{Z}_2$ lattice gauge theory. The ansatzes are designed to respect local and global symmetries of the original Hamiltonian and therefore act on a finely segmented state Hilbert space. Following Larocca et al. (2023), we numerically study the dimension of the dynamical Lie algebra (DLA) and the rank of the quantum Fisher information matrix (QFIM) of the ansatzes within specific invariant subspaces. The ansatzes all involve sums of weight-three Paulis in their generators, which is a feature that have so far been underexplored in this context. We also perform numerical experiments to determine the ground state energy of the original Hamiltonian via variational quantum eigensolver (VQE), and observe that overparametrization of the ansatzes coincides with the apparent disappearance of local minima in the loss function, in line with the finding in the reference. Finally, the decay rate of the VQE loss function under gradient descent optimization is revealed to scale linearly with the number of parameters in the ansatz. These results help to enrich the theory of overparameterization of quantum circuits and inform the design of scalable variational ansatzes.

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 paper numerically examines five symmetry-respecting Hamiltonian variational ansatzes (HVAs) constructed from weight-3 Pauli generators for the (1+1)d Z_2 lattice gauge theory. It computes the dimension of the dynamical Lie algebra (DLA) and the rank of the quantum Fisher information matrix (QFIM) inside invariant subspaces, then runs VQE experiments showing that overparametrization (via DLA dimension and QFIM rank) coincides with the apparent disappearance of local minima, while the gradient-descent decay rate of the loss scales linearly with the number of variational parameters.

Significance. If the reported linear scaling of optimization rate with parameter count holds beyond the specific model and ansatz family, the results would usefully inform the construction of scalable variational circuits for gauge-theory simulations. The work supplies concrete numerical data for a previously underexplored class of weight-3 generators and reproduces the overparametrization phenomenology of Larocca et al. (2023) in a gauge-theory setting; these are modest but genuine contributions to the empirical literature on quantum circuit overparametrization.

major comments (2)
  1. [Numerical VQE experiments] The central claim that the VQE loss decay rate under gradient descent scales linearly with parameter count rests entirely on numerical runs for five specific HVAs; no analytic derivation or scaling argument is supplied, and the manuscript provides no cross-model or cross-generator ablation (e.g., varying Pauli weight or spacetime dimension) that would test whether the linearity is an artifact of the chosen generators or the low-dimensional invariant subspaces.
  2. [Abstract and § on VQE results] The abstract and the description of the VQE runs do not report error bars, number of independent trials, shot-noise controls, or statistical tests for the linear scaling or for the disappearance of local minima; without these, the quantitative support for both empirical observations remains only moderate.
minor comments (2)
  1. [Ansatz construction] Notation for the five ansatzes (e.g., how the symmetry-invariant subspaces are segmented) should be introduced earlier and used consistently when reporting DLA dimensions and QFIM ranks.
  2. [Methods] The reference to Larocca et al. (2023) is appropriate, but a brief recap of the precise DLA/QFIM definitions employed would improve readability for readers outside the immediate subfield.

Simulated Author's Rebuttal

2 responses · 2 unresolved

We thank the referee for the careful review and constructive comments on our manuscript. We address each major comment below, indicating where revisions will be incorporated and providing honest clarifications on the numerical scope of the work.

read point-by-point responses

  1. Referee: [Numerical VQE experiments] The central claim that the VQE loss decay rate under gradient descent scales linearly with parameter count rests entirely on numerical runs for five specific HVAs; no analytic derivation or scaling argument is supplied, and the manuscript provides no cross-model or cross-generator ablation (e.g., varying Pauli weight or spacetime dimension) that would test whether the linearity is an artifact of the chosen generators or the low-dimensional invariant subspaces.

    Authors: Our work is a numerical study focused on five symmetry-preserving HVAs for the specific (1+1)d Z2 lattice gauge theory using weight-3 generators. The linear scaling is observed consistently in the VQE experiments across these ansatzes. No analytic derivation is supplied because the primary contribution lies in the numerical exploration of overparametrization phenomenology in this gauge-theory setting, reproducing and extending Larocca et al. (2023). We did not perform cross-model ablations, which would be a natural extension but lies outside the manuscript's scope. We will revise to explicitly note the numerical character of the observation and list the absence of broader ablations as a limitation. revision: partial

  2. Referee: [Abstract and § on VQE results] The abstract and the description of the VQE runs do not report error bars, number of independent trials, shot-noise controls, or statistical tests for the linear scaling or for the disappearance of local minima; without these, the quantitative support for both empirical observations remains only moderate.

    Authors: We agree that reporting error bars, the number of independent trials, shot-noise considerations, and any statistical measures would improve the quantitative presentation. This information is available from our simulations but was omitted. We will revise the abstract and the VQE results section to include these details. revision: yes

standing simulated objections not resolved
  • Analytic derivation or scaling argument for the linear scaling of the VQE loss decay rate with parameter count
  • Cross-model or cross-generator ablations to test the generality of the observed linearity

Circularity Check

0 steps flagged

No circularity; all central claims rest on direct numerical experiments and external reference

full rationale

The paper performs numerical studies of DLA dimension, QFIM rank, VQE loss landscapes, and gradient-descent decay rates for five specific HVAs on the (1+1)d Z2 model. It explicitly follows the numerical protocol of Larocca et al. (2023) and reports observations 'in line with the finding in the reference.' No equation or result is shown to reduce by construction to a parameter fitted from the same data, no self-citation supplies a uniqueness theorem or ansatz, and no derivation is claimed to be first-principles. The linear scaling is presented as an empirical observation, not an analytic prediction derived from the paper's own definitions. This is the normal case of a self-contained numerical study.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The work rests on standard quantum-computing assumptions plus the modeling choice that the chosen generators capture the relevant symmetry sectors; no new free parameters or invented entities are introduced.

axioms (2)
  • domain assumption The variational ansatzes are constructed to respect the local and global symmetries of the target Hamiltonian
    Explicitly stated as the design principle that segments the Hilbert space.
  • domain assumption Numerical VQE runs on finite systems faithfully indicate the absence of local minima in the infinite-parameter limit
    Implicit in the interpretation of the optimization landscapes.

pith-pipeline@v0.9.1-grok · 5806 in / 1367 out tokens · 37412 ms · 2026-06-28T01:32:06.937640+00:00 · methodology

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

45 extracted references · 1 canonical work pages · 1 internal anchor

  1. [1]

    The ansatz unitaries in each family are elements of a representation of a subgroup ofSU(2 2Ns)

    Consequently, it is also an eigenstate ofT 2 andCP witht= 0 andr= 0. The ansatz unitaries in each family are elements of a representation of a subgroup ofSU(2 2Ns). This represen- tation is reducible for all five families of our HVA, as is evident from the existence of invariants (quantum num- bers) discussed in the previous section, and our choice of HVA...

    2000

  2. [2]

    Larocca, N

    M. Larocca, N. Ju, D. Garc´ ıa-Mart´ ın, P. J. Coles, and M. Cerezo, Theory of overparametrization in quan- tum neural networks, Nat Comput Sci3, 542 (2023), arXiv:2109.11676 [quant-ph]

    arXiv 2023

  3. [3]

    J. R. McClean, S. Boixo, V. N. Smelyanskiy, R. Babbush, and H. Neven, Barren plateaus in quantum neural net- work training landscapes, Nat Commun9, 4812 (2018), arXiv:1803.11173 [quant-ph]

    Pith/arXiv arXiv 2018

  4. [4]

    E. R. Anschuetz, Critical Points in Quantum Generative Models (2021), arXiv:2109.06957 [quant-ph]

    arXiv 2021

  5. [5]

    Rivera-Dean, P

    J. Rivera-Dean, P. Huembeli, A. Ac´ ın, and J. Bowles, Avoiding local minima in Variational Quantum Algo- rithms with Neural Networks (2021), arXiv:2104.02955 [quant-ph]

    arXiv 2021

  6. [6]

    Wierichs, C

    D. Wierichs, C. Gogolin, and M. Kastoryano, Avoiding local minima in variational quantum eigensolvers with the natural gradient optimizer, Phys Rev Res2, 043246 (2020), arXiv:2004.14666 [quant-ph]

    arXiv 2020

  7. [7]

    You and X

    X. You and X. Wu, Exponentially many local minima in quantum neural networks, inProceedings of the 38th International Conference on Machine Learning, Proceed- ings of Machine Learning Research, Vol. 139, edited by M. Meila and T. Zhang (PMLR, 2021) pp. 12144–12155, arXiv:2110.02479 [quant-ph]

    arXiv 2021

  8. [8]

    E. R. Anschuetz and B. T. Kiani, Quantum variational algorithms are swamped with traps, Nat Commun13, 7760 (2022), arXiv:2205.05786 [quant-ph]

    arXiv 2022

  9. [9]

    Larocca, P

    M. Larocca, P. Czarnik, K. Sharma, G. Muraleedharan, P. J. Coles, and M. Cerezo, Diagnosing Barren Plateaus with Tools from Quantum Optimal Control, Quantum6, 824 (2022), arXiv:2105.14377 [quant-ph]

    arXiv 2022

  10. [10]

    M. E. S. Morales, J. D. Biamonte, and Z. Zimbor´ as, On the universality of the quantum approximate optimiza- tion algorithm, Quantum Inf Process19, 291 (2020), arXiv:1909.03123 [quant-ph]

    arXiv 2020

  11. [11]

    S. Kazi, M. Larocca, M. Farinati, P. J. Coles, M. Cerezo, and R. Zeier, Analyzing the quantum approximate opti- mization algorithm: Ans¨ atze, symmetries, and lie alge- bras, PRX Quantum6, 040345 (2025)

    2025

  12. [12]

    R. Mao, P. Yuan, J. Allcock, and S. Zhang, QAOA- MaxCut has barren plateaus for almost all graphs (2025), arXiv:2512.24577 [quant-ph]

    arXiv 2025

  13. [13]

    Matos, C

    G. Matos, C. N. Self, Z. Papi´ c, K. Meichanetzidis, and H. Dreyer, Characterization of variational quantum algo- rithms using free fermions, Quantum7, 966 (2023)

    2023

  14. [14]

    Allcock, M

    J. Allcock, M. Santha, P. Yuan, and S. Zhang, On the dy- namical Lie algebras of quantum approximate optimiza- tion algorithms (2024), arXiv:2407.12587 [quant-ph]

    Pith/arXiv arXiv 2024

  15. [15]

    Kordonowy and H

    S. Kordonowy and H. Leipold, The Lie algebra of XY- mixer topologies and warm starting QAOA for con- strained optimization, npj Quantum Inf12, 61 (2026)

    2026

  16. [16]

    D’Alessandro and Y

    D. D’Alessandro and Y. Isik, Controllability of the periodic quantum Ising spin chain and the Onsager algebra, J Phys A: Math Theor58, 115202 (2025), arXiv:2405.00898 [quant-ph]

    arXiv 2025

  17. [17]

    Tsvelikhovskiy, B

    B. Tsvelikhovskiy, B. Bach, J. Falla, and I. Safro, Reductions of qaoa induced by classical symmetries: Theoretical insights and practical implications (2026), arXiv:2602.16141 [quant-ph]

    Pith/arXiv arXiv 2026

  18. [18]

    Monbroussou, E

    L. Monbroussou, E. Z. Mamon, J. Landman, A. B. Grilo, R. Kukla, and E. Kashefi, Trainability and ex- pressivity of hamming-weight preserving quantum cir- cuits for machine learning, Quantum9, 1745 (2025), arXiv:2309.15547 [quant-ph]

    arXiv 2025

  19. [19]

    Wiersema, E

    R. Wiersema, E. K¨ okc¨ u, A. F. Kemper, and B. N. Bakalov, Classification of dynamical Lie algebras of 2- local spin systems on linear, circular and fully connected topologies, npj Quantum Inf.10, 110 (2024)

    2024

  20. [20]

    K¨ okc¨ u, R

    E. K¨ okc¨ u, R. Wiersema, A. F. Kemper, and B. N. Bakalov, Classification of dynamical lie algebras gener- ated by spin interactions on undirected graphs (2024)

    2024

  21. [21]

    Lloyd, Quantum approximate optimization is compu- tationally universal (2018), arXiv:1812.11075 [quant-ph]

    S. Lloyd, Quantum approximate optimization is compu- tationally universal (2018), arXiv:1812.11075 [quant-ph]

    Pith/arXiv arXiv 2018

  22. [22]

    Controllability of Symmetric Spin Networks

    F. Albertini and D. D’Alessandro, Controllability of symmetric spin networks, J Math Phys59, 10.1063/1.5004652 (2018), arXiv:1803.06689 [quant-ph]

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1063/1.5004652 2018

  23. [23]

    S. Kazi, M. Larocca, and M. Cerezo, On the universality ofS n-equivariantk-body gates, New J Phys26, 053030 (2024), arXiv:2303.00728 [quant-ph]

    arXiv 2024

  24. [24]

    Schatzki, M

    L. Schatzki, M. Larocca, Q. T. Nguyen, F. Sauvage, and M. Cerezo, Theoretical guarantees for permutation- equivariant quantum neural networks, npj Quantum Inf 10 10, 12 (2024), arXiv:2210.09974 [quant-ph]

    arXiv 2024

  25. [25]

    Adamatti Bridi, D

    G. Adamatti Bridi, D. Lim, L. Pira, R. Azevedo Medeiros Santos, F. de Lima Marquezino, and S. Adhikary, Expressivity limits in quantum walk- based optimization, Phys Rev A113, 022423 (2026), arXiv:2508.05749 [quant-ph]

    arXiv 2026

  26. [26]

    Tsvelikhovskiy, M

    B. Tsvelikhovskiy, M. Nuyten, and B. N. Bakalov, Prov- able avoidance of barren plateaus for the Quantum Ap- proximate Optimization Algorithm with Grover mixers (2025), arXiv:2509.10424 [quant-ph]

    arXiv 2025

  27. [27]

    Aguilar, S

    G. Aguilar, S. Cichy, J. Eisert, and L. Bittel, Full clas- sification of Pauli Lie algebras (2024), arXiv:2408.00081 [quant-ph]

    arXiv 2024

  28. [28]

    F. Azad, M. Inajetovic, S. K¨ uhn, and A. Pappa, Barren- plateau free variational quantum simulation of Z2 lattice gauge theories (2025), arXiv:2507.19203 [quant-ph]

    arXiv 2025

  29. [29]

    Mazzola, S

    G. Mazzola, S. V. Mathis, G. Mazzola, and I. Tavernelli, Gauge-invariant quantum circuits forU(1) and Yang- Mills lattice gauge theories, Phys Rev Res3, 043209 (2021), arXiv:2105.05870 [quant-ph]

    arXiv 2021

  30. [30]

    Alexandrou, A

    C. Alexandrou, A. Athenodorou, K. Blekos, G. Polykratis, and S. K¨ uhn, Realizing string breaking dynamics in aZ 2 lattice gauge theory on quantum hard- ware, Phys Rev D112, 114506 (2025), arXiv:2504.13760 [hep-lat]

    arXiv 2025

  31. [31]

    Kogut and L

    J. Kogut and L. Susskind, Hamiltonian formulation of Wilson’s lattice gauge theories, Phys Rev D11, 395 (1975)

    1975

  32. [32]

    Mildenberger, W

    J. Mildenberger, W. Mruczkiewicz, J. C. Halimeh, Z. Jiang, and P. Hauke, Confinement in aZ2 lattice gauge theory on a quantum computer, Nat Phys21, 312 (2025), arXiv:2203.08905 [quant-ph]

    arXiv 2025

  33. [33]

    Charles, E

    C. Charles, E. J. Gustafson, E. Hardt, F. Herren, N. Hogan, H. Lamm, S. Starecheski, R. S. Van de Water, and M. L. Wagman, SimulatingZ 2 lattice gauge theory on a quantum computer, Phys Rev E109, 015307 (2024), arXiv:2305.02361 [hep-lat]

    arXiv 2024

  34. [34]

    Bringewatt, J

    J. Bringewatt, J. Kunjummen, and N. Mueller, Random- ized measurement protocols for lattice gauge theories, Quantum8, 1300 (2024), arXiv:2303.15519 [quant-ph]

    arXiv 2024

  35. [35]

    Borla, R

    U. Borla, R. Verresen, F. Grusdt, and S. Moroz, Con- fined phases of one-dimensional spinless fermions coupled toz 2 gauge theory, Phys Rev Lett124, 120503 (2020), arXiv:1909.07399 [quant-ph]

    arXiv 2020

  36. [36]

    Desaules, T

    J.-Y. Desaules, T. Iadecola, and J. C. Halimeh, Mass- assisted local deconfinement in a confinedZ 2 lat- tice gauge theory, Phys Rev B112, 014301 (2025), arXiv:2404.11645 [cond-mat.quant-gas]

    arXiv 2025

  37. [37]

    Wecker, M

    D. Wecker, M. B. Hastings, and M. Troyer, Progress towards practical quantum variational algorithms, Phys Rev A92, 042303 (2015), arXiv:1507.08969 [quant-ph]

    Pith/arXiv arXiv 2015

  38. [38]

    W. W. Ho and T. H. Hsieh, Efficient variational simula- tion of non-trivial quantum states, SciPost Phys6, 029 (2019), arXiv:1803.00026 [cont-mat.str-el]

    Pith/arXiv arXiv 2019

  39. [39]

    Wiersema, C

    R. Wiersema, C. Zhou, Y. de Sereville, J. F. Car- rasquilla, Y. B. Kim, and H. Yuen, Exploring en- tanglement and optimization within the hamiltonian variational ansatz, PRX Quantum1, 020319 (2020), arXiv:2008.02941 [quant-ph]

    arXiv 2020

  40. [40]

    J. Liu, K. Najafi, K. Sharma, F. Tacchino, L. Jiang, and A. Mezzacapo, Analytic theory for the dynamics of wide quantum neural networks, Phys Rev Lett130, 150601 (2023), arXiv:2203.16711 [quant-ph]

    arXiv 2023

  41. [41]

    S. G. Schirmer, H. Fu, and A. I. Solomon, Complete con- trollability of quantum systems, Phys Rev A63, 063410 (2001), arXiv:quant-ph/0010031 [quant-ph]

    Pith/arXiv arXiv 2001

  42. [42]

    H. R. Grimsley, S. E. Economou, E. Barnes, and N. J. Mayhall, An adaptive variational algorithm for exact molecular simulations on a quantum computer, Nat Com- mun10, 3007 (2019), arXiv:1812.11173 [quant-ph]

    Pith/arXiv arXiv 2019

  43. [43]

    R. C. Farrell, M. Illa, A. N. Ciavarella, and M. J. Sav- age, Scalable circuits for preparing ground states on digital quantum computers: The schwinger model vac- uum on 100 qubits, PRX Quantum5, 020315 (2024), arXiv:2308.04481 [quant-ph]

    arXiv 2024

  44. [44]

    Gross, K

    D. Gross, K. Audenaert, and J. Eisert, Evenly distributed unitaries: On the structure of unitary designs, J Math Phys48, 052104 (2007), arXiv:quant-ph/0611002 [quant- ph]

    Pith/arXiv arXiv 2007

  45. [45]

    Nakata, Y

    Y. Nakata, Y. Takeuchi, M. Kliesch, and A. Dar- mawan, Computational complexity of unitary and state design properties, PRX Quantum6, 030345 (2025), arXiv:2410.23353 [quant-ph]. Appendix A: Bound on the rank of the QFIM from the number of degrees of freedom of the ansatz The proof of Eq. (26) is as follows. Let S= span{|ψ ⟨M⟩(θ)⟩ |θ∈R M , M∈N}(A1) be the ...

    arXiv 2025