pith. sign in

arxiv: 2606.25600 · v1 · pith:LQR4BQA6new · submitted 2026-06-24 · 🪐 quant-ph · cond-mat.dis-nn· cs.LG· physics.comp-ph

Two-dimensional Hyperbolic RNN Neural Quantum State

H. L. Dao This is my paper

Pith reviewed 2026-06-25 21:19 UTC · model grok-4.3

classification 🪐 quant-ph cond-mat.dis-nncs.LGphysics.comp-ph
keywords hyperbolic neural quantum statesLorentz RNN2D transverse field Ising modelconformal field theoryAdS/CFT dualityphase transitionvariational Monte Carlorecurrent neural networks
0
0 comments X

The pith

Hyperbolic Lorentz 2DRNN neural quantum states outperform Euclidean 2DRNN at the critical point of the 2D transverse-field Ising model.

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

This paper builds the first two-dimensional hyperbolic neural quantum state using a Lorentz 2DRNN ansatz. It tests this against the standard Euclidean 2DRNN on the 2D transverse field Ising model for lattices up to size 12. The hyperbolic version performs better specifically at the phase transition, where the system's physics is a conformal field theory dual to an Anti-de-Sitter space with hyperbolic geometry. The work also shows that one-dimensional hyperbolic NQS improve on Euclidean versions when the 2D lattice is treated as a 1D chain with added hierarchical interactions.

Core claim

The Lorentz 2DRNN neural quantum state in two dimensions outperforms its Euclidean counterpart in the N by N 2D transverse field Ising model exactly at the critical transverse field, where the ground state is described by a conformal field theory whose holographic dual is an Anti-de-Sitter space with hyperbolic spatial geometry. This outperformance is presented as evidence for the utility of hyperbolic geometries in neural quantum states for systems at criticality. The same pattern holds when one-dimensional hyperbolic RNN and GRU networks are applied to the 2D model after mapping it to one dimension.

What carries the argument

The Lorentz 2DRNN, a two-dimensional recurrent neural network whose parameters and states are defined in the Lorentz model of hyperbolic geometry, used as the variational wavefunction ansatz for the neural quantum state.

Load-bearing premise

The performance difference is caused by the hyperbolic geometry corresponding to the AdS dual of the CFT rather than by unrelated differences in how the networks are constructed or trained.

What would settle it

Measuring the relative performance of hyperbolic and Euclidean 2DRNN NQS in the same model but at transverse field values away from the critical point, or performing controlled experiments that equalize network expressivity.

Figures

Figures reproduced from arXiv: 2606.25600 by H. L. Dao.

Figure 1
Figure 1. Figure 1: Schematic of the process of calculating the RNN wavefunction Ψ( [PITH_FULL_IMAGE:figures/full_fig_p006_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: A comparison of the performances of Euclidean 2DRNN and Lorentz 2DRNN (both best and averaged [PITH_FULL_IMAGE:figures/full_fig_p011_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: The spin-spin correlation decay length curves versus distance [PITH_FULL_IMAGE:figures/full_fig_p012_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: The individual spin-spin correlation decay length curves for Lorentz 2DRNN and Euclidean 2D [PITH_FULL_IMAGE:figures/full_fig_p013_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: The individual spin-spin correlation decay length curves for Lorentz 2DRNN (with [PITH_FULL_IMAGE:figures/full_fig_p013_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: A comparison of the performances of Euclidean 2DRNN and Lorentz 2DRNN in the VMC experiments [PITH_FULL_IMAGE:figures/full_fig_p015_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: The best performances of 1D Euclidean and hyperbolic Poincar´e/Lorentz NQS ansatzes, chosen from [PITH_FULL_IMAGE:figures/full_fig_p021_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: The performances of 1D Euclidean and hyperbolic Poincar´e/Lorentz NQS ansatzes, averaged over [PITH_FULL_IMAGE:figures/full_fig_p022_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: The ⟨σ z 0σ z r ⟩ spin-spin correlation decay length curves for all six types of 1D NQS ansatzes from the 2D TFIM with (N, N) = (8, 8). From top to bottom: In the top subfigure, the curves for all six ansatzes are shown together, in the middle subfigure, the zoomed-in curves of the three RNN-based NQS variants are shown, in the bottom subfigure, the zoomed-in curves of the three GRU-based NQS variants are … view at source ↗
Figure 10
Figure 10. Figure 10: The ⟨σ z 0σ z r ⟩ spin-spin correlation decay length curves for all six types of 1D NQS ansatzes from the 2D TFIM with (N, N) = (10, 10). From top to bottom: In the top subfigure, the curves for all six ansatzes are shown together, in the middle subfigure, the zoomed-in curves of the three RNN-based NQS variants are shown, in the bottom subfigure, the zoomed-in curves of the three GRU-based NQS variants a… view at source ↗
Figure 11
Figure 11. Figure 11: The ⟨σ z 0σ z r ⟩ spin-spin correlation decay length curves for all six types of 1D NQS ansatzes from the 2D TFIM with (N, N) = (12, 12). From top to bottom: In the top subfigure, the curves for all six ansatzes are shown together, in the middle subfigure, the zoomed-in curves of the three RNN-based NQS variants are shown, in the bottom subfigure, the zoomed-in curves of the three GRU-based NQS variants a… view at source ↗
Figure 12
Figure 12. Figure 12: An illustration of the interaction hierarchies of different degrees of nearest neighbor interactions [PITH_FULL_IMAGE:figures/full_fig_p028_12.png] view at source ↗
Figure 13
Figure 13. Figure 13: An illustration of the interaction hierarchies of different degrees of nearest neighbor interactions in [PITH_FULL_IMAGE:figures/full_fig_p029_13.png] view at source ↗
read the original abstract

In the first part of this work, we construct the first type of two-dimensional (2D) hyperbolic neural quantum state (NQS) in the form of the Lorentz 2DRNN (Recurrent Neural Network) and benchmark its performance against the Euclidean 2DRNN in the paradigmatic $N\times N$ 2D Transverse Field Ising Model (2DTFIM) setting with different lattice sizes up to $N=12$ and at different transverse magnetic field strengths. We find that hyperbolic Lorentz 2DRNN NQS definitively outperform Euclidean 2DRNN NQS when the system is at the phase transition point when the physics can be described by a conformal field theory (CFT), which is known to be dual to an Anti-de-Sitter (AdS) space whose spatial geometry is hyperbolic. In the second part of this work, we benchmark the performances of the recently introduced one-dimensional (1D) hyperbolic NQS including Poincar\'e RNN/GRU and Lorentz RNN/GRU against their Euclidean NQS versions in $N\times N$ 2DTFIM, which has to be converted to a one-dimensional setting to allow for the use of 1D NQS. The findings in this case extend our previous results that 1D hyperbolic NQS definitively outperform 1D Euclidean NQS, thanks to the combined effects of the hierarchical structure comprising the first and $N^{th}$ neighbor interactions present in the 1D system arising from the 2D lattice and the CFT physics at the critical point. While more studies with larger system sizes are required, our work serves as a proof-of-concept for the utility, effectiveness as well as the superior performances of one- and two-dimensional hyperbolic NQS ansatzes compared to the existing Euclidean NQS in many-body quantum physics systems, especially when these systems exhibit structural hierarchy or when they are at criticality, or a combination of both.

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

3 major / 1 minor

Summary. The manuscript constructs the first 2D hyperbolic neural quantum state ansatz in the form of a Lorentz 2DRNN and benchmarks it against the Euclidean 2DRNN on the N×N 2D transverse-field Ising model (2DTFIM) for lattices up to N=12 at multiple transverse-field values. It reports that the hyperbolic version outperforms the Euclidean version at the critical point, attributing the advantage to the match between hyperbolic geometry and the AdS space dual to the CFT description of the critical theory. The work also benchmarks 1D hyperbolic NQS (Poincaré and Lorentz RNN/GRU) against Euclidean versions on the 2D model mapped to 1D, extending prior 1D results and linking gains to hierarchical neighbor interactions plus criticality.

Significance. If the performance advantage can be shown to arise specifically from the hyperbolic geometry rather than other factors, the result would supply a concrete proof-of-concept that AdS/CFT-inspired geometries can improve NQS expressivity for critical many-body systems. The explicit acknowledgment that larger lattices are required and the direct comparison on the same Hamiltonian are positive features, but the current evidence base remains too preliminary for strong claims about holographic utility in quantum simulation.

major comments (3)
  1. [Abstract] Abstract: the claim that hyperbolic Lorentz 2DRNN NQS 'definitively outperform' Euclidean 2DRNN NQS at the phase transition supplies no numerical metrics, error bars, training protocols, or statistical tests, so the magnitude and reliability of the reported advantage cannot be assessed.
  2. [Benchmark comparisons] Benchmark section (2D Lorentz 2DRNN vs. Euclidean 2DRNN): no information is given on parameter counts, hidden-state dimensions, optimizer settings, initialization, or any ablation that holds all other architectural elements fixed while toggling only the hyperbolic versus Euclidean metric; without these controls the attribution of the gap to AdS/CFT geometry matching remains unsupported.
  3. [1D hyperbolic NQS extension] 1D hyperbolic NQS benchmarks on the mapped 2D system: the attribution of improvement to 'the combined effects of the hierarchical structure comprising the first and Nth neighbor interactions ... and the CFT physics' likewise lacks controlled experiments that isolate geometry from expressivity or optimization differences.
minor comments (1)
  1. [Abstract] Abstract: the phrasing 'when the physics can be described by a conformal field theory (CFT)' would benefit from an explicit statement of the critical transverse-field value used for the 2DTFIM.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for the constructive feedback, which highlights opportunities to strengthen the presentation of our results and experimental controls. We address each major comment below and will revise the manuscript to improve clarity, add missing details, and qualify claims where appropriate.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the claim that hyperbolic Lorentz 2DRNN NQS 'definitively outperform' Euclidean 2DRNN NQS at the phase transition supplies no numerical metrics, error bars, training protocols, or statistical tests, so the magnitude and reliability of the reported advantage cannot be assessed.

    Authors: We agree that the abstract should be more quantitative to allow assessment of the advantage. In the revised version we will replace the phrasing 'definitively outperform' with a qualified statement that includes representative metrics (e.g., relative energy error reductions at criticality), mention of error bars obtained from multiple independent runs, and a brief reference to the shared training protocol (Adam optimizer, fixed hidden dimension, same number of epochs). This will be done without altering the underlying data or conclusions. revision: yes

  2. Referee: [Benchmark comparisons] Benchmark section (2D Lorentz 2DRNN vs. Euclidean 2DRNN): no information is given on parameter counts, hidden-state dimensions, optimizer settings, initialization, or any ablation that holds all other architectural elements fixed while toggling only the hyperbolic versus Euclidean metric; without these controls the attribution of the gap to AdS/CFT geometry matching remains unsupported.

    Authors: We will add a new table (or expanded methods subsection) that reports, for both models: hidden-state dimension, total trainable parameters, optimizer (Adam with learning-rate schedule), weight initialization, batch size, and number of training steps. The architectures are identical except for the metric tensor used in the recurrent update; we will explicitly state this and discuss why the comparison isolates the geometric contribution. We acknowledge that a full factorial ablation isolating every possible factor is not present and will qualify the geometric interpretation accordingly while noting that the observed gap is reproducible across the reported lattice sizes. revision: partial

  3. Referee: [1D hyperbolic NQS extension] 1D hyperbolic NQS benchmarks on the mapped 2D system: the attribution of improvement to 'the combined effects of the hierarchical structure comprising the first and Nth neighbor interactions ... and the CFT physics' likewise lacks controlled experiments that isolate geometry from expressivity or optimization differences.

    Authors: We will expand the corresponding methods paragraph to list the identical hyper-parameters (hidden dimension, optimizer settings, initialization) used for the Poincaré/Lorentz and Euclidean 1D models. We will clarify that the hierarchical neighbor structure is an unavoidable consequence of the 2D-to-1D mapping and that performance differences appear most pronounced at the critical point. The attribution will be rephrased as an interpretation consistent with the data rather than a definitive causal claim, and we will note the absence of exhaustive ablations that would fully disentangle geometry from other factors. revision: partial

Circularity Check

0 steps flagged

No circularity: empirical benchmarks only

full rationale

The manuscript reports direct empirical performance comparisons between Lorentz 2DRNN NQS and Euclidean 2DRNN NQS (and 1D variants) on the 2DTFIM Hamiltonian across lattice sizes up to N=12 and varying transverse fields. The central finding—that hyperbolic versions outperform at the critical point—is presented as an observed benchmark result rather than a derived prediction or first-principles claim. No equations, fitted parameters, or self-citations are invoked in a manner that reduces any reported advantage to an input defined by the authors themselves. The work is therefore self-contained against external benchmarks with no load-bearing circular steps.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

Review is abstract-only; no explicit free parameters, invented entities, or detailed axioms are stated. The central claim rests on the domain assumption that NQS performance differences can be attributed to geometry matching CFT/AdS duality.

axioms (2)
  • domain assumption Recurrent neural networks can serve as variational ansatzes for quantum many-body wavefunctions
    Implicit foundation of all NQS work referenced in the abstract
  • domain assumption The 2DTFIM at criticality is described by a CFT whose spatial geometry is dual to hyperbolic AdS space
    Invoked to explain why hyperbolic NQS should outperform Euclidean ones

pith-pipeline@v0.9.1-grok · 5892 in / 1426 out tokens · 26529 ms · 2026-06-25T21:19:48.080979+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

49 extracted references · 1 canonical work pages

  1. [1]

    Carleo and M

    G. Carleo and M. Troyer, Solving the Quantum Many-Body Problem with Artificial Neural Networks, Science 355, 602 (2017), arXiv:1606.02318 [cond-mat.dis-nn]

    Pith/arXiv arXiv 2017

  2. [2]

    Huang and L

    L. Huang and L. Wang, Accelerate Monte Carlo Simulations with Restricted Boltzmann Machines, arXiv:1610.02746v2

    Pith/arXiv arXiv

  3. [3]

    Cai and J

    Z. Cai and J. Liu, Approximating quantum many-body wave-functions using artificial neural networks,Phys. Rev. B 97, 035116 (2018), arXiv:1704.05148 [cond-mat.str-el]

    Pith/arXiv arXiv 2018

  4. [4]

    Saito and M

    H. Saito and M. Kato, Machine learning technique to find quantum many-body ground states of bosons on a lattice, J. Phys. Soc. Jpn. 87, 014001 (2018), arXiv:1709.05468 [cond-mat.dis-nn]

    Pith/arXiv arXiv 2018

  5. [5]

    X. Liang, Wen-Yuan Liu, Pei-Ze Lin, Guang-Can Guo, Yong-Sheng Zhang, and Lixin He, Solving frustrated quantum many-particle models with convolutional neural networks, arXiv: 1807.09422v2

    Pith/arXiv arXiv

  6. [6]

    C. Roth, A. Szab´ o, and A. H. MacDonald, High-accuracy variational Monte Carlo for frustrated magnets with deep neural networks, Phys. Rev. B 108, 054410 (2023), arXiv:2211.07749v2 [cond-mat.str-el]

    arXiv 2023

  7. [7]

    Hibat-Allah, M

    M. Hibat-Allah, M. Ganahl, L. E. Hayward, R. G. Melko, and J. Carrasquill, Recurrent neural network wave functions, Physical Review Research 2, 023358 (2020)

    2020

  8. [8]

    Hibat-Allah, R

    M. Hibat-Allah, R. G. Melko, J. Carrasquilla, Supplementing Recurrent Neural Network Wave Functions with Symmetry and Annealing to Improve Accuracy, Machine Learning and the Physical Sciences, NeurIPS 2021, arXiv:2207.14314v2 [cond-mat.dis-nn]

    arXiv 2021

  9. [9]

    Hibat-Allah, E

    M. Hibat-Allah, E. Merali, G. Torlai, R. G. Melko and J. Carrasquilla, Recurrent neural network wave functions for Rydberg atom arrays on kagome lattice, arXiv:2405.20384v1 [cond-mat.quant-gas]

    arXiv

  10. [10]

    Sprague and S

    K. Sprague and S. Czischek, Variational Monte Carlo with Large Patched Transformers, Commun Phys 7, 90 (2024), arXiv:2306.03921 [quant-ph]

    arXiv 2024

  11. [11]

    Lange, G

    H. Lange, G. Bornet, G. Emperauger, C. Chen, T. Lahaye, S. Kienle, A. Browaeys, A. Bohrdt, Transformer neural networks and quantum simulators: a hybrid approach for simulating strongly correlated systems, Quantum 9, 1675 (2025), arXiv:2406.00091 [cond-mat.dis-nn]

    arXiv 2025

  12. [12]

    H. L. Dao, Hyperbolic recurrent neural network as the first type of non-Euclidean neural quantum state ansatz, Eur. Phys. J. Plus 141:199, arXiv:2505.22083 [quant-ph, cond-mat.dis-nn, cs.LG, physics.comp-ph] (2026)

    arXiv 2026

  13. [13]

    H. L. Dao, New non-Euclidean neural quantum state ansatzes from additional types of hyperbolic neural networks, arXiv:2604.2337 [quant-ph, cs.LG, cond-mat.dis-nn]

    arXiv

  14. [14]

    Ganea, G

    O.-E. Ganea, G. Becigneul, and T. Hofmann, Hyperbolic Neural Networks, Advances in Neural Information Processing Systems 31, pages 5345–5355. Curran Associates, Inc. arXiv: 1805.09112 [cs.LG]

    Pith/arXiv arXiv

  15. [15]

    N. He, M. Yang and Rex Ying, HyperCore: The Core Framework for Building Hyperbolic Foundation Models with Comprehensive Modules, arXiv:2504.08912 [cs.LG]

    arXiv

  16. [16]

    Becca \ and\ author S

    F. Becca and S. Sorella, Quantum Monte Carlo approaches for correlated systems, Cambridge University Press 2017, DOI: 10.1017/9781316417041

    work page doi:10.1017/9781316417041 2017

  17. [17]

    Chung, C

    J. Chung, C. Gulcehre, K. Cho, Y. Bengio, Gated feedback recurrent neural networks, in: ICML, 2015

    2015

  18. [18]

    H. L. Dao, Deep Learning Calabi-Yau four folds with hybrid and recurrent neural network architectures, Nucl. Phys. B 1013 (2025) 116832, arXiv:2405.17406 [hep-th, cs.LG, math.AG]

    arXiv 2025

  19. [19]

    W. Chen, X. Han, Y. Lin, H. Zhao, Z. Liu, P. Li, M. Sun, J. Zhou, Fully Hyperbolic Neural Networks, in ACL 2022 Main Conference, arXiv:2105.14686 [cs.CL]

    arXiv 2022

  20. [20]

    W. Peng, T. Varanka, A. Mostafa, H. Shi, G. Zhao, Hyperbolic Deep Neural Networks: A Survey, arXiv:2101.04562 [cs.LG, cs.CV]

    arXiv

  21. [21]

    Gulcehre, M

    C. Gulcehre, M. Denil, M. Malinowski, A. Razavi, R. Pascanu, et. al., Hyperbolic Attention Networks, arXiv:1805.09786v1 [cs.NE]

    Pith/arXiv arXiv

  22. [22]

    Strube, A Fully Hyperbolic Neural Model for Hierarchical Multi-Class Classification, Findings of EMNLP2020, arXiv:2010.02053 [cs.CL]

    F.Lopez and M. Strube, A Fully Hyperbolic Neural Model for Hierarchical Multi-Class Classification, Findings of EMNLP2020, arXiv:2010.02053 [cs.CL]. 32

    arXiv 2010

  23. [23]

    Shimizu, Y

    R. Shimizu, Y. Mukuta and T. Harada, Hyperbolic Neural Networks++, The Ninth International Confer- ence on Learning Representations (ICLR 2021), arXiv:2006.08210 [cs.LG]

    arXiv 2021

  24. [24]

    Mathieu, C

    E. Mathieu, C. Le Lan, C. J. Maddison, R. Tomioka, and Y. W. Teh, Continuous Hierarchical Represen- tations with Poincar´ e Variational Auto-Encoders, arXiv:1901.06033

    arXiv 1901

  25. [25]

    Bachmann, G

    G. Bachmann, G. Becigneul, O.-E. Ganea, Constant Curvature Graph Convolutional Networks, arXiv:1911.05076v3

    arXiv 1911

  26. [26]

    Linial, E

    N. Linial, E. London, and Y. Rabinovich. The geometry of graphs and some of its algorithmic applications, Combinatorica, 15(2):215–245, 1995

    1995

  27. [27]

    Krioukov, F

    D. Krioukov, F. Papadopoulos, A. Vahdat, and M. Bogun´ a, Curvature and temperature of complex net- works, Physical Review E, 80(3):035101, 2009

    2009

  28. [28]

    Krioukov, F

    D. Krioukov, F. Papadopoulos, M. Kitsak, A. Vahdat, and M. Bogun´ a, Hyperbolic geometry of complex networks, Physical Review E, 82(3):036106, 2010

    2010

  29. [29]

    H. W. J. Bl¨ ote and Y. Deng, Cluster Monte Carlo simulation of the transverse Ising model, Phys. Rev. E 66, 066110 (2002)

    2002

  30. [30]

    R. Sarkar. Low distortion Delaunay embedding of trees in hyperbolic plane, In Proc. of the International Symposium on Graph Drawing (GD 2011), pages 355–366, Eindhoven, Netherlands, 2011

    2011

  31. [31]

    F. Sala, C. De Sa, A. Gu, and C. R´ e. 2018. Representation tradeoffs for hyperbolic embeddings. In Inter- national Conference on Machine Learning, pages 4457–4466

    2018

  32. [32]

    Bonnabel, Stochastic gradient descent on riemannian manifolds, IEEE Transactions on Automatic Con- trol, 58(9):2217–2229, Sept 2013

    S. Bonnabel, Stochastic gradient descent on riemannian manifolds, IEEE Transactions on Automatic Con- trol, 58(9):2217–2229, Sept 2013

    2013

  33. [33]

    Ganea, G

    O.-E. Ganea, G. B´ ecigneul, and T. Hofmann. Hyperbolic entailment cones for learning hierarchical embed- dings. In Proceedings of the thirty-fifth international conference on machine learning (ICML), 2018

    2018

  34. [34]

    B´ ecigneul and O.-E Ganea, Riemannian Adaptive Optimization Methods, International Conference on Learning Representations (ICLR) (2019), arXiv:1810.00760 [cs.LG]

    G. B´ ecigneul and O.-E Ganea, Riemannian Adaptive Optimization Methods, International Conference on Learning Representations (ICLR) (2019), arXiv:1810.00760 [cs.LG]

    Pith/arXiv arXiv 2019

  35. [35]

    Nickel and D

    M. Nickel and D. Kiela, Poincar´ e embeddings for learning hierarchical representations, arXiv:1705.08039 [cs.AI]

    Pith/arXiv arXiv

  36. [36]

    Nickel and D

    M. Nickel and D. Kiela, Learning Continuous Hierarchies, in the Lorentz Model of Hyperbolic Geometry, ICML 2018, arXiv:1806.03417 [cs.AI]

    Pith/arXiv arXiv 2018

  37. [37]

    H. L. Dao, Exploring new variational quantum circuit ansatzes for solvingSU(2) matrix models, Eur. Phys. J. C (2025) 85:705, arXiv:2503.13368 [quant-ph, hep-th, physics.comp-ph]

    arXiv 2025

  38. [38]

    El-Showk, M

    S. El-Showk, M. F. Paulos, D. Poland, S. Rychkov, D. Simmons-Duffin, A. Vichi, Solving the 3d Ising Model with the Conformal Bootstrap II. c-Minimization and Precise Critical Exponents, J. Stat. Phys. 157, 869-914 (2014), arXiv:1403.4545 [hep-th]

    Pith/arXiv arXiv 2014

  39. [39]

    Vidal, Entanglement renormalization, Phys

    G. Vidal, Entanglement renormalization, Phys. Rev. Lett. 99, 220405 (2007), arXiv:cond-mat/0512165 [cond-mat.str-el]

    Pith/arXiv arXiv 2007

  40. [40]

    Vidal, A class of quantum many-body states that can be efficiently simulated, Phys

    G. Vidal, A class of quantum many-body states that can be efficiently simulated, Phys. Rev. Lett. 101, 110501 (2008) (2007), arXiv:quant-ph/0610099

    Pith/arXiv arXiv 2008

  41. [41]

    Evenbly and G

    G. Evenbly and G. Vidal, Entanglement renormalization in two spatial dimensions, Phys. Rev. Lett. 102, 180406 (2009), arXiv:0811.0879 [cond-mat.str-el]

    Pith/arXiv arXiv 2009

  42. [42]

    Swingle, Entanglement Renormalization and Holography, Phys

    B. Swingle, Entanglement Renormalization and Holography, Phys. Rev. D 86, 065007 (2012), arXiv:0905.1317 [cond-mat.str-el]

    Pith/arXiv arXiv 2012

  43. [43]

    Swingle, Constructing holographic spacetimes using entanglement renormalization, arXiv:1209.3304 [hep-th]

    B. Swingle, Constructing holographic spacetimes using entanglement renormalization, arXiv:1209.3304 [hep-th]

    Pith/arXiv arXiv

  44. [44]

    Rinaldi, X

    E. Rinaldi, X. Han, M. Hassan, Y. Feng, F. Nori, M. McGuigan, M. Hanada, Matrix-model simulations using quantum computing, deep learning, and lattice Monte Carlo. PRX Quantum 3, 010324 (2022)

    2022

  45. [45]

    J. M. Maldacena, The large N limit of superconformal field theories and supergravity, Adv. Theor. Math. Phys. 2 (1998) 231 [Int. J. Theor. Phys. 38 (1999) 1113] [arXiv:hep- th/9711200]. 33

    arXiv 1998

  46. [46]

    Di Francesco, P

    P. Di Francesco, P. Mathieu, D. S´ en´ echal, Conformal Field Theory, Springer-Verlag (1997)

    1997

  47. [47]

    Sachdev, Quantum Phase Transitions, Cambridge University Press (2011)

    S. Sachdev, Quantum Phase Transitions, Cambridge University Press (2011)

    2011

  48. [48]

    D. J. Watts and S. H. Strogatz. Collective dynamics of ‘small-world’ networks. Nature, 393(6684):440, 1998

    1998

  49. [49]

    Nickel and D

    Q.Liu, M. Nickel and D. Kiela, Hyperbolic Graph Neural Networks, arxiv:1910.12892 [cs.LG] 34

    arXiv 1910