pith. sign in

arxiv: 2605.18725 · v1 · pith:WKWAA2LPnew · submitted 2026-05-18 · ✦ hep-th

Signatures of Quantum Chaos in the D1D5 System

Haoyu Zhang This is my paper

Pith reviewed 2026-05-20 09:01 UTC · model grok-4.3

classification ✦ hep-th
keywords D1D5 CFTquantum chaosrandom-matrix statisticslevel repulsionnear-BPS sectorsfinite-N effectsorbifold CFTnon-planar mixing
0
0 comments X

The pith

Non-planar mixing at finite N produces random-matrix level statistics in the D1D5 CFT

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

The paper investigates the emergence of random-matrix statistics in the D1D5 CFT through second-order lifting matrices in low-energy near-BPS sectors. It compares the finite-N case with N=3 to the planar large-N limit at fixed orbifold conformal weight and R charges. In the planar limit, mixing between single-cycle and multi-cycle states is suppressed, leading to Poisson-like level statistics in symmetry-resolved sectors. At finite N, non-planar terms enable this mixing, which is accompanied by level repulsion consistent with random-matrix behavior. This matters because it suggests that non-planar cycle-structure mixing is linked to the onset of quantum chaos in accessible low-energy sectors of the theory.

Core claim

In the planar large-N limit at fixed orbifold energy, mixing between single-cycle and multi-cycle states is suppressed, and the symmetry-resolved lifting spectra display Poisson-like level statistics. At finite N, non-planar terms restore this mixing between different cycle structures. Within the resulting symmetry-resolved sectors, this finite-N mixing is accompanied by level repulsion consistent with random-matrix behavior. These results suggest that non-planar cycle-structure mixing at finite N is associated with the onset of level repulsion and random-matrix-like spacing statistics in the low-energy near-BPS sectors.

What carries the argument

Second-order lifting matrices in symmetry-resolved sectors that encode the finite-N mixing between different cycle structures in the D1D5 CFT.

If this is right

  • In the planar large-N limit mixing is suppressed and symmetry-resolved spectra follow Poisson statistics.
  • Finite-N non-planar corrections restore mixing between single- and multi-cycle states.
  • The restored mixing produces level repulsion that matches random-matrix expectations.
  • The transition appears specifically in low-energy near-BPS sectors at fixed orbifold weight and R-charges.

Where Pith is reading between the lines

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

  • The same finite-N mixing mechanism could be checked in other orbifold CFTs to test whether chaos onset is a general feature.
  • Extending the analysis to higher-order lifts might reveal whether repulsion strengthens or saturates at smaller N.
  • If these sectors correspond to black-hole microstates, the mixing-chaos link could constrain how chaos emerges in the dual geometry.

Load-bearing premise

The second-order lifting matrices computed in the low-energy near-BPS sectors fully capture the relevant mixing and level statistics, and the symmetry-resolved sectors chosen for analysis are representative of the onset of random-matrix behavior.

What would settle it

A direct computation showing Poisson statistics persisting at finite N despite non-planar mixing, or random-matrix repulsion appearing without cycle-structure mixing in the low-energy sectors.

Figures

Figures reproduced from arXiv: 2605.18725 by Haoyu Zhang.

Figure 1
Figure 1. Figure 1: FIG. 1. Nearest-neighbor spacing distributions at ( [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
read the original abstract

We investigate the emergence of random-matrix statistics in the D1D5 CFT by studying second-order lifting matrices in low-energy near-BPS sectors. We compare the $N=3$ finite-$N$ lifting problems with the planar large-$N$ limit at fixed orbifold conformal weight and R charges. In the planar large-$N$ limit at fixed orbifold energy, mixing between single-cycle and multi-cycle states is suppressed, and the symmetry-resolved lifting spectra display Poisson-like level statistics. At finite $N$, non-planar terms restore this mixing between different cycle structures. Within the resulting symmetry-resolved sectors, this finite-$N$ mixing is accompanied by level repulsion consistent with random-matrix behavior. These results suggest that, in the low-energy near-BPS sectors accessible to our analysis, non-planar cycle-structure mixing at finite $N$ is associated with the onset of level repulsion and random-matrix-like spacing statistics.

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 investigates the emergence of random-matrix statistics in the D1D5 orbifold CFT by computing second-order lifting matrices in low-energy near-BPS sectors. It directly compares the N=3 finite-N lifting problem with the planar large-N limit at fixed orbifold conformal weight and R-charges. In the large-N planar limit, mixing between single-cycle and multi-cycle states is suppressed, yielding Poisson-like level statistics in symmetry-resolved sectors. At finite N, non-planar terms restore cycle-structure mixing, which is accompanied by level repulsion consistent with random-matrix behavior. The authors conclude that non-planar mixing at finite N drives the onset of random-matrix-like spacing statistics in these accessible sectors.

Significance. If the central claim holds after addressing statistical limitations, the work provides a concrete, calculable example linking finite-N non-planar effects to the onset of quantum chaos in a holographic CFT. The direct comparison of two regimes of the same model (planar vs. finite-N) without external fitting parameters is a strength, offering potential insight into how non-planar corrections induce level repulsion in near-BPS states of the D1D5 system, with broader relevance to AdS3/CFT2 and the emergence of thermalization.

major comments (2)
  1. [§4 (finite-N lifting spectra and level statistics)] The central claim rests on level-repulsion signatures in symmetry-resolved sectors at N=3. However, these sectors have small Hilbert-space dimensions, yielding few eigenvalues for spacing analysis after unfolding. With such limited statistics, apparent repulsion can arise from finite-size effects or incomplete unfolding rather than genuine random-matrix behavior induced by the mixing; quantitative measures (e.g., Brody parameter with uncertainties or comparison to larger ensembles) are needed to substantiate the claim.
  2. [§3.2 (lifting matrices) and §5 (discussion of sectors)] The analysis assumes that the computed second-order lifting matrices fully capture the relevant mixing and that the chosen symmetry-resolved sectors are representative. Higher-order corrections or different sector selections could alter the observed spacing distributions, directly affecting the association between non-planar mixing and random-matrix statistics.
minor comments (2)
  1. [§4] The unfolding procedure applied to the spectra prior to computing spacing distributions should be described explicitly, including any binning or normalization choices, to ensure reproducibility of the Poisson vs. RMT comparison.
  2. [§2] Notation for orbifold conformal weight and R-charges is used without a dedicated summary table; adding one would clarify the fixed quantities in the large-N vs. finite-N comparison.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the careful reading and constructive comments on our manuscript. We address the major concerns point by point below, indicating where revisions have been made to strengthen the presentation and analysis.

read point-by-point responses

  1. Referee: [§4 (finite-N lifting spectra and level statistics)] The central claim rests on level-repulsion signatures in symmetry-resolved sectors at N=3. However, these sectors have small Hilbert-space dimensions, yielding few eigenvalues for spacing analysis after unfolding. With such limited statistics, apparent repulsion can arise from finite-size effects or incomplete unfolding rather than genuine random-matrix behavior induced by the mixing; quantitative measures (e.g., Brody parameter with uncertainties or comparison to larger ensembles) are needed to substantiate the claim.

    Authors: We agree that the symmetry-resolved sectors at N=3 have small dimensions, limiting the number of eigenvalues available after unfolding and making the statistics inherently modest. This is an unavoidable feature of exact computations at small finite N. In the revised manuscript we have added the Brody parameter (with uncertainties obtained by bootstrap resampling of the available eigenvalues) to §4 as a quantitative measure of repulsion, together with an explicit comparison of the unfolded spacings to both Poisson and GOE distributions. We have also clarified the unfolding procedure. While larger-N ensembles would be desirable, the direct planar-versus-finite-N comparison within the same model already isolates the effect of non-planar mixing; we therefore regard the added diagnostics as sufficient to support the central claim at the level of the accessible data. revision: partial

  2. Referee: [§3.2 (lifting matrices) and §5 (discussion of sectors)] The analysis assumes that the computed second-order lifting matrices fully capture the relevant mixing and that the chosen symmetry-resolved sectors are representative. Higher-order corrections or different sector selections could alter the observed spacing distributions, directly affecting the association between non-planar mixing and random-matrix statistics.

    Authors: Second-order degenerate perturbation theory supplies the leading correction in the near-BPS regime we study; higher-order terms are suppressed by additional powers of the effective coupling and are therefore parametrically small. In the revised §5 we have added an explicit discussion of this suppression and of why we expect the qualitative level-repulsion signature to survive. We have also verified that the observed repulsion is robust across several representative sectors that exhibit non-planar cycle mixing at fixed orbifold weight and R-charges, and we have included a brief note on this robustness. We therefore maintain that the chosen sectors are representative of the low-energy physics under consideration. revision: yes

Circularity Check

0 steps flagged

Direct computation of lifting matrices and spectra yields Poisson-to-RMT contrast with no circularity

full rationale

The derivation proceeds by explicit construction and diagonalization of second-order lifting matrices in two regimes (planar large-N at fixed orbifold weight, and finite N=3). The resulting eigenvalues are then subjected to standard spacing analysis within symmetry-resolved sectors. Both the suppression of mixing (and Poisson statistics) in the large-N case and the restoration of mixing (and apparent level repulsion) at finite N follow directly from the matrix elements computed in each regime; neither statistic is fitted to the target distribution nor defined in terms of the other. The paper remains self-contained against external benchmarks, with no load-bearing self-citation, ansatz smuggling, or self-definitional step.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

Based solely on the abstract; no explicit free parameters, new entities, or ad-hoc axioms are stated. The work relies on standard orbifold CFT techniques whose details are not supplied here.

axioms (1)
  • domain assumption The D1D5 system is described by an orbifold sigma-model CFT whose second-order lifting matrices can be computed in near-BPS sectors.
    The entire analysis presupposes the validity of this standard framework for the D1D5 CFT.

pith-pipeline@v0.9.0 · 5678 in / 1384 out tokens · 45443 ms · 2026-05-20T09:01:48.251285+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

66 extracted references · 66 canonical work pages · 14 internal anchors

  1. [1]

    In both the finite-Nand planar large-Ncomputations, the lifting matrices are determined up to an overall universal nor- malization factor.5 The first sectors in our data set with enough levels to display qualitative spectral-statistical behavior occur ath= 4 andh= 9

    work page

  2. [2]

    The result- ing comparison between the planar large-Nand finite-N spacing distributions is shown in Fig

    For the sectors displayed below, the relevantR-charge isj= 0 whenh is integer andj= 1 2 whenhis half-integer. The result- ing comparison between the planar large-Nand finite-N spacing distributions is shown in Fig. 1. In both exam- ples, the planar large-Nspectra retain substantial weight near small spacings and are closer to Poisson statistics, while the...

    work page

  3. [3]

    This agrees with the sector decomposition above, since the entry 2 3 in the sector (2, 3

    Therefore the eigenvalue 2 3, inherited from theh= 1 primary spectrum, is expected to appear with total multiplicity 2×3 = 6. This agrees with the sector decomposition above, since the entry 2 3 in the sector (2, 3

    work page

  4. [4]

    Forh= 2 andj= 0, the lifting spectrum decomposes by Casimir sector as Casimir sector Lifting eigenvalues (0,0) 1, 3 8 ( 3 4 , 3 4) 2× 1 2 (2,2) 2 3 ( 15 4 , 3

    contributes 1×3×2 = 6 states. Forh= 2 andj= 0, the lifting spectrum decomposes by Casimir sector as Casimir sector Lifting eigenvalues (0,0) 1, 3 8 ( 3 4 , 3 4) 2× 1 2 (2,2) 2 3 ( 15 4 , 3

    work page

  5. [5]

    For example, a primary at h= 1,j= 0 has five global descendants ath= 2,j= 0

    {1} (2,0) 2× 2 3 Ath= 2, the descendants of theh= 1,j= 0 primaries give another consistency check. For example, a primary at h= 1,j= 0 has five global descendants ath= 2,j= 0. Since there are three primary states with lifting eigenvalue 2 3 ath= 1, we expect 5×3 = 15 descendants with the same lifting eigenvalue ath= 2. This agrees with the sector-by-secto...

    work page

  6. [6]

    Since there are six such primaries ath= 1, we expect 2×6 = 12 corresponding descendants ath= 3 2,j= 1

    In particular, the eigenvalue 2 3 appearing ath= 3 2,j= 1 2 can be identified with the global descendants of theh= 1,j= 0 primaries with the same lifting eigenvalue. Since there are six such primaries ath= 1, we expect 2×6 = 12 corresponding descendants ath= 3 2,j= 1

    work page

  7. [7]

    The same reasoning applies to the inherited eigenvalue 16

    This agrees with the multiplicity 3 + 3×3 = 12 found in the sector decomposition above. The same reasoning applies to the inherited eigenvalue 16

    work page

  8. [8]

    Forh= 2 and j= 0, one finds Casimir sector Lifting eigenvalues (0, 3 4) 1, 3 8 ,2× 8 27 ,2× 32 243 ( 3 4 ,0) 2× 1 2 , 32 81 ,0.3078. . . ,0.2598. . . ,0.2148. . . ,2× 16 81 ,0.1279. . . ,0.0865. . . , ( 3 4 ,2) 2× 1 2 , 32 81 , 16 81 (2, 3 4) 3× 2 3 , 32 81 ,0.3098. . . , 64 243 ,0.2219. . . ,2× 40 243 (2, 15 4 ) 2 3 ( 15 4 ,0) 1,2× 8 27 ( 15 4 ,2) 1,2× 8...

    work page

  9. [9]

    Since each such primary has five global descendants ath= 2,j= 0, we expect the eigenvalue 2 3 to appear with total multiplicity 5×6 = 30 ath= 2,j= 0

    Ath= 1,j= 0, this eigenvalue is associated with six primary states. Since each such primary has five global descendants ath= 2,j= 0, we expect the eigenvalue 2 3 to appear with total multiplicity 5×6 = 30 ath= 2,j= 0. This agrees with the sector-by-sector decomposition. The eigenvalue 2 3 appears with multiplicity 3 in the sector (2, 3 4), and with multip...

    work page

  10. [10]

    Planar (2,0),(0,2),(0,0) (h, j) = ( 9 2 , 1

    work page

  11. [11]

    N= 3 (0,0),(0,2),(2,0),(2,2) 10 These are the sectors whose unfolded spacings are pooled in the corresponding histograms in Fig. 1. The spectra in these sectors are provided in the ancillary filedata.tex

    work page

  12. [12]

    P. Saad, S. H. Shenker, and D. Stanford, (2019), arXiv:1903.11115 [hep-th]

    work page internal anchor Pith review Pith/arXiv arXiv 2019

  13. [13]

    Stanford and E

    D. Stanford and E. Witten, Adv. Theor. Math. Phys.24, 1475 (2020), arXiv:1907.03363 [hep-th]

    work page arXiv 2020

  14. [14]

    Witten, Proc

    E. Witten, Proc. Roy. Soc. Lond. A476, 20200582 (2020), arXiv:2006.13414 [hep-th]

    work page arXiv 2020

  15. [15]

    Maxfield and G

    H. Maxfield and G. J. Turiaci, JHEP01, 118 (2021), arXiv:2006.11317 [hep-th]

    work page arXiv 2021

  16. [16]

    T. G. Mertens and G. J. Turiaci, Living Rev. Rel.26, 4 (2023), arXiv:2210.10846 [hep-th]

    work page arXiv 2023

  17. [17]

    G. J. Turiaci and E. Witten, JHEP12, 003 (2023), arXiv:2305.19438 [hep-th]

    work page arXiv 2023

  18. [18]

    Cotler and K

    J. Cotler and K. Jensen, JHEP04, 033 (2021), arXiv:2006.08648 [hep-th]

    work page arXiv 2021

  19. [19]

    Cotler and K

    J. Cotler and K. Jensen, JHEP11, 058 (2020), arXiv:2007.15653 [hep-th]

    work page arXiv 2020

  20. [20]

    Belin, J

    A. Belin, J. de Boer, D. L. Jafferis, P. Nayak, and J. Sonner, JHEP09, 163 (2024), arXiv:2308.03829 [hep-th]

    work page arXiv 2024

  21. [21]

    D. L. Jafferis, L. Rozenberg, and D. Wang, JHEP10, 228 (2025), arXiv:2506.19817 [hep-th]

    work page arXiv 2025

  22. [22]

    D. L. Jafferis, L. Rozenberg, D. Sarkar, and D. Wang, (2025), arXiv:2512.05045 [hep-th]

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  23. [23]

    D. L. Jafferis, L. Rozenberg, and G. Wong, JHEP02, 208 (2025), arXiv:2407.02649 [hep-th]

    work page arXiv 2025

  24. [24]

    Schlenker and E

    J.-M. Schlenker and E. Witten, JHEP07, 143 (2022), arXiv:2202.01372 [hep-th]

    work page arXiv 2022

  25. [25]

    Cotler and K

    J. Cotler and K. Jensen, JHEP11, 070 (2022), arXiv:2205.12968 [hep-th]

    work page arXiv 2022

  26. [26]

    McLoughlin, R

    T. McLoughlin, R. Pereira, and A. Spiering, SciPost Phys.14, 049 (2023), arXiv:2011.04633 [hep-th]

    work page arXiv 2023

  27. [27]

    McLoughlin and A

    T. McLoughlin and A. Spiering, JHEP09, 240 (2022), arXiv:2202.12075 [hep-th]

    work page arXiv 2022

  28. [28]

    Dijkgraaf, G

    R. Dijkgraaf, G. W. Moore, E. P. Verlinde, and H. L. Verlinde, Commun. Math. Phys.185, 197 (1997), arXiv:hep- th/9608096

    work page arXiv 1997

  29. [29]

    J. R. David, G. Mandal, and S. R. Wadia, Phys. Rept.369, 549 (2002), arXiv:hep-th/0203048

    work page internal anchor Pith review Pith/arXiv arXiv 2002

  30. [30]

    Proving the PP-Wave/CFT_2 Duality

    E. Gava and K. S. Narain, JHEP12, 023 (2002), arXiv:hep-th/0208081

    work page internal anchor Pith review Pith/arXiv arXiv 2002

  31. [31]

    S. G. Avery, B. D. Chowdhury, and S. D. Mathur, JHEP06, 031 (2010), arXiv:1002.3132 [hep-th]

    work page internal anchor Pith review Pith/arXiv arXiv 2010

  32. [32]

    S. G. Avery, B. D. Chowdhury, and S. D. Mathur, JHEP06, 032 (2010), arXiv:1003.2746 [hep-th]

    work page internal anchor Pith review Pith/arXiv arXiv 2010

  33. [33]

    Guo and S

    B. Guo and S. D. Mathur, JHEP03, 028 (2020), arXiv:1912.05567 [hep-th]

    work page arXiv 2020

  34. [34]

    Guo and S

    B. Guo and S. D. Mathur, JHEP11, 145 (2020), arXiv:2008.01274 [hep-th]

    work page arXiv 2020

  35. [35]

    M. R. Gaberdiel, R. Gopakumar, and B. Nairz, JHEP06, 030 (2024), arXiv:2312.13288 [hep-th]

    work page arXiv 2024

  36. [36]

    M. V. Berry and M. Tabor, Proc. R. Soc. Lond. A356, 375 (1977)

    work page 1977

  37. [37]

    Bohigas, M

    O. Bohigas, M. J. Giannoni, and C. Schmit, Phys. Rev. Lett.52, 1 (1984)

    work page 1984

  38. [38]

    T. Guhr, A. M¨ uller-Groeling, and H. A. Weidenm¨ uller, Phys. Rept.299, 189 (1998), arXiv:cond-mat/9707301

    work page internal anchor Pith review Pith/arXiv arXiv 1998

  39. [39]

    Haake,Quantum Signatures of Chaos, 3rd ed

    F. Haake,Quantum Signatures of Chaos, 3rd ed. (Springer, Berlin, 2010)

    work page 2010

  40. [40]

    Frolov and A

    S. Frolov and A. Sfondrini, JHEP08, 179 (2024), arXiv:2312.14114 [hep-th]

    work page arXiv 2024

  41. [41]

    M. R. Gaberdiel, F. Lichtner, and B. Nairz, (2024), arXiv:2411.17612 [hep-th]

    work page arXiv 2024

  42. [42]

    M. R. Gaberdiel, B. Nairz, and C. Peng, (2026), arXiv:2605.06465 [hep-th]

    work page internal anchor Pith review Pith/arXiv arXiv 2026

  43. [43]

    Chang, Y.-H

    C.-M. Chang, Y.-H. Lin, and H. Zhang, (2025), arXiv:2501.05448 [hep-th]

    work page arXiv 2025

  44. [44]

    Chang and H

    C.-M. Chang and H. Zhang, (2025), arXiv:2511.23294 [hep-th]

    work page arXiv 2025

  45. [45]

    M. L. Mehta,Random Matrices, 3rd ed., Pure and Applied Mathematics, Vol. 142 (Elsevier/Academic Press, Amsterdam, 2004)

    work page 2004

  46. [46]

    Correlation functions for M^N/S_N orbifolds

    O. Lunin and S. D. Mathur, Commun. Math. Phys.219, 399 (2001), arXiv:hep-th/0006196

    work page internal anchor Pith review Pith/arXiv arXiv 2001

  47. [47]

    Three-point functions for M^N/S^N orbifolds with N=4 supersymmetry

    O. Lunin and S. D. Mathur, Commun. Math. Phys.227, 385 (2002), arXiv:hep-th/0103169

    work page internal anchor Pith review Pith/arXiv arXiv 2002

  48. [48]

    Dei and L

    A. Dei and L. Eberhardt, JHEP01, 108 (2020), arXiv:1911.08485 [hep-th]

    work page arXiv 2020

  49. [49]

    Fabri, A

    M. Fabri, A. Sfondrini, and T. Skrzypek, (2025), arXiv:2504.13091 [hep-th]

    work page arXiv 2025

  50. [50]

    Budzik, H

    K. Budzik, H. Murali, and P. Vieira, (2023), arXiv:2306.04693 [hep-th]

    work page arXiv 2023

  51. [51]

    Chang, Y

    C.-M. Chang, Y. Chen, B. S. Sia, and Z. Yang, (2024), arXiv:2412.06902 [hep-th]

    work page arXiv 2024

  52. [52]

    Microscopic Origin of the Bekenstein-Hawking Entropy

    A. Strominger and C. Vafa, Phys. Lett. B379, 99 (1996), arXiv:hep-th/9601029

    work page internal anchor Pith review Pith/arXiv arXiv 1996

  53. [53]

    Lectures on black holes and the AdS_3 / CFT_2 correspondence

    P. Kraus, Lect. Notes Phys.755, 193 (2008), arXiv:hep-th/0609074

    work page internal anchor Pith review Pith/arXiv arXiv 2008

  54. [54]

    Chang and Y.-H

    C.-M. Chang and Y.-H. Lin, (2024), arXiv:2402.10129 [hep-th]

    work page arXiv 2024

  55. [55]

    Choi and E

    J. Choi and E. Lee, (2025), arXiv:2511.09519 [hep-th]

    work page arXiv 2025

  56. [56]

    S. Kim, J. Lee, S. Lee, and H. Oh, (2025), arXiv:2511.03105 [hep-th]

    work page arXiv 2025

  57. [57]

    Belin, P

    A. Belin, P. Singh, R. Vadala, and A. Zaffaroni, (2025), arXiv:2512.04146 [hep-th]

    work page arXiv 2025

  58. [58]

    M. R. R. Hughes and M. Shigemori, (2025), arXiv:2505.14888 [hep-th]

    work page arXiv 2025

  59. [59]

    Chen, (2025), arXiv:2511.00790 [hep-th]

    Y. Chen, (2025), arXiv:2511.00790 [hep-th]

    work page arXiv 2025

  60. [60]

    H. W. Lin, J. Maldacena, L. Rozenberg, and J. Shan, SciPost Phys.14, 150 (2023), arXiv:2207.00407 [hep-th]

    work page arXiv 2023

  61. [61]

    H. W. Lin, J. Maldacena, L. Rozenberg, and J. Shan, SciPost Phys.14, 128 (2023), arXiv:2207.00408 [hep-th]

    work page arXiv 2023

  62. [62]

    Y. Chen, H. W. Lin, and S. H. Shenker, (2024), arXiv:2407.19387 [hep-th]

    work page arXiv 2024

  63. [63]

    Y. Chen, S. Colin-Ellerin, O. Mamroud, and K. Papadodimas, (2026), arXiv:2604.23287 [hep-th]

    work page internal anchor Pith review Pith/arXiv arXiv 2026

  64. [64]

    Eguchi and A

    T. Eguchi and A. Taormina, Phys. Lett. B196, 75 (1987)

    work page 1987

  65. [65]

    Eguchi and A

    T. Eguchi and A. Taormina, Phys. Lett. B200, 315 (1988)

    work page 1988

  66. [66]

    M. R. R. Hughes, K. Jin, D. Matsumoto, L. Miyahara, and M. Shigemori, (2026), arXiv:2604.20663 [hep-th]

    work page internal anchor Pith review Pith/arXiv arXiv 2026