Pith. sign in

REVIEW 4 minor 274 references

Classic random-matrix spectral laws can be derived transparently with the cavity method and then applied directly to nuclear spectra, ecosystem stability, PCA, and localisation.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · grok-4.5

2026-07-10 16:21 UTC pith:VUHAX35E

load-bearing objection Solid, cavity-first lecture notes that re-derive the classic RMT laws and tools cleanly; useful reference, not a research claim.

arxiv 2607.07868 v1 pith:VUHAX35E submitted 2026-07-08 cond-mat.dis-nn

Lecture notes on random matrix theory: the results, the applications, and the analytical tools

Joseph W. Baron This is my paper
classification cond-mat.dis-nn MSC 60B2015B5282B44 PACS 05.40.-a02.10.Yn05.45.Mt
keywords random matrix theorycavity methodWigner semicircleelliptic lawMarchenko-PasturAnderson localisationdynamic mean-field theoryfree probability
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

These lecture notes show that the best-known results of random matrix theory—the semicircle law, the elliptic law, the Marchenko-Pastur law, the Kesten-McKay law, and the associated spacing statistics—can be obtained in a self-contained way by the cavity (block-inversion) method. Each law is then linked to a concrete application: nuclear energy-level spacings, May’s stability criterion for large ecosystems, principal-component analysis and the BBP transition, Anderson localisation on random regular graphs, and the spectra of complex networks. A second part re-derives the same results with the diagrammatic, replica, path-integral and supersymmetric formalisms, free probability and population dynamics, so that a practitioner can see when each tool is most useful. The notes therefore give both a pedagogical route into the classic theorems and a practical reference for choosing the right analytic method.

Core claim

In the large-N limit, under standard moment conditions on the matrix entries, the cavity equations for the resolvent close and concentrate, yielding the classic spectral densities (semicircle, elliptic, Marchenko-Pastur, Kesten-McKay, …) and the associated outlier eigenvalues and spacing statistics; the same densities control the listed applications once the appropriate random matrix is identified.

What carries the argument

The cavity (Schur-complement / block-inversion) method applied to the resolvent: deleting one row and column produces self-consistent equations for the diagonal Green functions that become deterministic by concentration of large sums, after which the Stieltjes inversion recovers the eigenvalue density.

Load-bearing premise

The large-N concentration arguments (central-limit behaviour of cavity sums, vanishing of off-diagonal resolvent entries, tree-like factorisation on sparse graphs) remain uniformly valid for every ensemble and application treated.

What would settle it

For any of the ensembles (GOE, elliptic, Wishart, random-regular adjacency, …) compute the empirical resolvent or eigenvalue histogram at moderate but increasing N and check whether the deviation from the predicted density or outlier location scales as claimed; a systematic O(1) discrepancy that does not vanish would falsify the concentration step.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • Nuclear level-spacing histograms should match the Wigner surmise once the spectrum is unfolded.
  • An ecosystem whose Jacobian has mean and variance exceeding the May thresholds will be unstable, with the nature of the instability (oscillatory versus exponential) fixed by whether the bulk ellipse or the outlier crosses the stability line.
  • PCA recovers a population spike only above the BBP threshold; below it the principal-component overlap vanishes and the sample spectrum is pure Marchenko-Pastur noise.
  • On a random regular graph the Anderson model exhibits a mobility edge separating extended bulk states from exponentially localised Lifshitz-tail states once the on-site disorder is large enough.
  • The same cavity equations supply the spectral edge that sets the epidemic threshold on a configuration-model network and the diffusion instability threshold of its Laplacian.

Where Pith is reading between the lines

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

  • Because the notes deliberately juxtapose cavity, replica, supersymmetry and free-probability derivations of the same semicircle, a reader can treat them as a controlled comparison of analytic cost versus range of validity for any new disordered model.
  • The explicit DMFT treatment of the spherical p-spin and generalised Lotka-Volterra systems suggests that the same cavity-plus-mean-field pipeline can be reused for other high-dimensional non-linear dynamics whose linearised Jacobians are random.
  • The population-dynamics algorithm of the final section is presented as a numerical solver for the cavity equations; it therefore offers a practical route to spectral densities of sparse or non-homogeneous ensembles that lack closed-form solutions.
  • Finance applications of Marchenko-Pastur cleaning are given only briefly, yet the rotationally-invariant estimator formulae are complete enough that a practitioner could implement optimal shrinkage on real covariance matrices without further derivation.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

0 major / 4 minor

Summary. These lecture notes derive the classic large-N spectral laws of random matrix theory (Wigner semicircle, Girko elliptic law and outliers, Marchenko-Pastur with BBP transition, Kesten-McKay, sparse-network spectra) primarily via the cavity/block-inversion method, then illustrate each law with a concrete application (nuclear level spacings, May ecosystems and Lotka-Volterra/neural stability, PCA and covariance cleaning, Anderson localisation on RRGs, network Laplacians). Part 2 re-derives the semicircle (and selected extensions) from first principles with diagrammatic, replica, supersymmetric and MSRJD path-integral methods, and adds free probability, population dynamics and Dyson Brownian motion. Exercises close most sections. The central claim is pedagogical: the cavity route is elementary and transparent under standard moment conditions, the same laws control the listed applications, and the advanced formalisms become useful in complementary regimes.

Significance. If the notes are adopted as a graduate reference they fill a genuine gap: a single, self-contained treatment that (i) derives the workhorse spectral laws with the cavity method rather than heavy combinatorics or replicas, (ii) immediately embeds each law in a modern application (ecology, finance, localisation, networks), and (iii) supplies a comparative toolkit of the advanced methods used in the disordered-systems literature. The derivations recover the known closed forms, numerical checks against single large matrices are shown throughout, and the exercises are well-chosen. The absence of free parameters or circular definitions, together with the explicit discussion of when each method is advantageous, makes the manuscript a high-value pedagogical resource for the cond-mat.dis-nn and adjacent communities.

minor comments (4)
  1. Throughout Part 1 the large-N concentration steps (CLT on cavity sums, vanishing of off-diagonal resolvent entries, tree-like factorisation) are invoked without quantitative error bounds or references to the rigorous literature that supplies them. A short paragraph or footnote in §II.D–F (and analogous places in §§III, V, VII) pointing to the relevant theorems would strengthen the notes without changing their pedagogical character.
  2. Figure captions and axis labels are occasionally terse (e.g. Figs. 7–11, 17–19). Adding the precise ensemble parameters and the meaning of solid/dashed curves would improve readability for students.
  3. A few typographical inconsistencies remain (Marčenko vs Marchenko, occasional missing spaces around equations). A light copy-edit pass would remove them.
  4. The population-dynamics section (XV) is very brief relative to the other advanced tools. A short worked example (e.g. the sparse ER cavity equations of §VIII) would make the method more immediately usable.

Circularity Check

0 steps flagged

No significant circularity: pedagogical re-derivations of classic RMT laws from block inversion, CLT concentration, and standard transforms, without fitted inputs or self-definitional loops.

full rationale

These are self-contained lecture notes. Part 1 derives the semicircle (II), elliptic law (III), DMFT order parameters (IV), Marchenko-Pastur/BBP (V), Coulomb-gas/DBM joint laws (VI), Kesten-McKay and localisation diagnostics (VII), and sparse-network spectra (VIII) by the cavity/block-inversion method plus large-N concentration of cavity sums and vanishing of off-diagonal resolvent entries. Each step starts from the definition of the resolvent (or Hermitised resolvent), applies the Schur complement, invokes CLT/moment conditions that are stated explicitly, and solves a closed equation for G(z) or C(z,z*); the density is then recovered by the inverse Stieltjes (or 2D) transform. No parameter is fitted to data and then re-presented as a prediction of that same data; numerical checks are comparisons, not calibration. Part 2 re-derives the same laws via diagrams, replicas, SUSY, path integrals, free probability, and population dynamics from first principles, again without circular definitions. Applications (nuclear spacings, May stability, PCA cleaning, Anderson IPR, etc.) use the derived spectra as inputs; they do not define the spectra in terms of the application outcomes. There is no load-bearing self-citation of an author-owned uniqueness theorem, no ansatz smuggled solely via self-citation, and no renaming of an empirical pattern as a new derivation. The concentration arguments lack quantitative error bounds, but that is a completeness/rigour issue, not circularity. Score 0 is therefore appropriate.

Axiom & Free-Parameter Ledger

0 free parameters · 3 axioms · 0 invented entities

As pure exposition the notes inherit the standard large-N assumptions of RMT (i.i.d. or weakly dependent entries with controlled moments, thermodynamic limit before spectral smoothing, tree-like factorisation on sparse graphs). No free parameters are fitted and no new entities are postulated.

axioms (3)
  • domain assumption Matrix entries are independent (up to Hermiticity) with vanishing mean, variance O(1/N) and higher moments o(1/N); the empirical spectral measure concentrates.
    Invoked throughout Part 1 (e.g. Eqs. (18), (40), (164)) to justify cavity-sum concentration and universality.
  • domain assumption Off-diagonal resolvent elements are negligible in the large-N limit.
    Used repeatedly (II.F, III.F, V.C) to close the cavity equations; standard but not proved with error bounds here.
  • domain assumption Sparse graphs are locally tree-like so that cavity factorisation holds.
    Central to the Anderson and network sections (VII.C, VIII).

pith-pipeline@v1.1.0-grok45 · 65362 in / 2061 out tokens · 28402 ms · 2026-07-10T16:21:17.872198+00:00 · methodology

0 comments
read the original abstract

Random matrix theory has established itself as a theoretical cornerstone of the mathematical sciences over the past century. It has undeniable utility in areas of research as diverse as nuclear physics, finance, ecology and disordered systems. The purpose of these notes is twofold. First, the most famous and widely used classic results are derived in a pedagogical manner, mostly using the comparatively elementary and transparent cavity method. The significance of each result is then demonstrated in the context of a particular application. There are also some select exercises at the end of each section. In the second part of these notes, a reference guide of analytical techniques for the random-matrix/disordered-systems practitioner is provided. Introducing the diagrammatic, replica, path-integral, and supersymmetric formalisms from first principles, we rederive some of the aforementioned classic results, particularly focussing on the simplest one -- the semicircle law. Innovations such as the population dynamics method and the tools of free probability theory are also included. We discuss the merits of each analytical approach, and we highlight the contexts in which each becomes particularly useful.

Figures

Figures reproduced from arXiv: 2607.07868 by Joseph W. Baron.

Figure 1
Figure 1. Figure 1: FIG. 1: The eigenvalues of a square matrix of size [PITH_FULL_IMAGE:figures/full_fig_p008_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2: A (vastly incomplete) timeline of random matrix theory. [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: FIG. 3: Histograms of the eigenvalues for a single symmetric Gaussian random matrix with [PITH_FULL_IMAGE:figures/full_fig_p012_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: FIG. 4: Comparison of Im[ [PITH_FULL_IMAGE:figures/full_fig_p018_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: FIG. 5: Universality of the semicircle law. As long as the elements [PITH_FULL_IMAGE:figures/full_fig_p020_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: FIG. 6: Comparison of numerical diagonalisation results to the Wigner surmise in Eq. (32) for [PITH_FULL_IMAGE:figures/full_fig_p023_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: FIG. 7: Example eigenvalue spectra of the matrix [PITH_FULL_IMAGE:figures/full_fig_p029_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: FIG. 8: Comparison of [PITH_FULL_IMAGE:figures/full_fig_p031_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: FIG. 9: Tests of the uniform eigenvalue density given in Eq. (63). (Left) Integrated eigenvalue [PITH_FULL_IMAGE:figures/full_fig_p036_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: FIG. 10: Sketch of [PITH_FULL_IMAGE:figures/full_fig_p038_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: FIG. 11: The nature of the dynamics depending on the spectrum. (Top) Stable fixed point with [PITH_FULL_IMAGE:figures/full_fig_p039_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: FIG. 12: Schematic illustration of the idea of mean-field theory. [PITH_FULL_IMAGE:figures/full_fig_p049_12.png] view at source ↗
Figure 13
Figure 13. Figure 13: FIG. 13: Relaxational dynamics of the system in Eq. (94) compared with the DMFT predictions [PITH_FULL_IMAGE:figures/full_fig_p057_13.png] view at source ↗
Figure 14
Figure 14. Figure 14: FIG. 14: Power spectrum of fluctuations for the system in Eq. (94) compared with the DMFT [PITH_FULL_IMAGE:figures/full_fig_p058_14.png] view at source ↗
Figure 15
Figure 15. Figure 15: FIG. 15: Solutions to Eq. (144) for various [PITH_FULL_IMAGE:figures/full_fig_p063_15.png] view at source ↗
Figure 16
Figure 16. Figure 16: FIG. 16: (Left): Some quasi-1-dimensional data. The location of a point in the ( [PITH_FULL_IMAGE:figures/full_fig_p068_16.png] view at source ↗
Figure 17
Figure 17. Figure 17: FIG. 17: Verification of the Marˇcenko-Pastur law and associated outlier in Eqs. (174) and (184). [PITH_FULL_IMAGE:figures/full_fig_p074_17.png] view at source ↗
Figure 18
Figure 18. Figure 18: FIG. 18: Comparison of numerical diagonalisation results to Eq. (194). Numerical results are for [PITH_FULL_IMAGE:figures/full_fig_p077_18.png] view at source ↗
Figure 19
Figure 19. Figure 19: FIG. 19: (Left) Comparison of the rescaled eigenvalues [PITH_FULL_IMAGE:figures/full_fig_p081_19.png] view at source ↗
Figure 20
Figure 20. Figure 20: FIG. 20: Eigenvalues that form the Wigner semicircle can be thought of as ‘charged particles’ [PITH_FULL_IMAGE:figures/full_fig_p089_20.png] view at source ↗
Figure 21
Figure 21. Figure 21: FIG. 21: Eigenvalue density for the potential [PITH_FULL_IMAGE:figures/full_fig_p092_21.png] view at source ↗
Figure 22
Figure 22. Figure 22: FIG. 22: Eigenvalue density correlations in the cases [PITH_FULL_IMAGE:figures/full_fig_p093_22.png] view at source ↗
Figure 23
Figure 23. Figure 23: FIG. 23: Illustration of the cavity factorisation on a tree-like graph. The highlighted node is [PITH_FULL_IMAGE:figures/full_fig_p101_23.png] view at source ↗
Figure 24
Figure 24. Figure 24: FIG. 24: Kesten-McKay law for [PITH_FULL_IMAGE:figures/full_fig_p103_24.png] view at source ↗
Figure 25
Figure 25. Figure 25: FIG. 25: The bulk and tail regions of the eigenvalue spectrum. Here, the prediction in Eq. (269) [PITH_FULL_IMAGE:figures/full_fig_p106_25.png] view at source ↗
Figure 26
Figure 26. Figure 26: FIG. 26: The eigenvalue density (numerical results represented by bars) and the IPR (red crosses [PITH_FULL_IMAGE:figures/full_fig_p107_26.png] view at source ↗
Figure 27
Figure 27. Figure 27: FIG. 27: Crossover from Wigner-Dyson statistics (in the continuous bulk part of the spectrum) [PITH_FULL_IMAGE:figures/full_fig_p110_27.png] view at source ↗
Figure 28
Figure 28. Figure 28: FIG. 28: Eigenvalue density for [PITH_FULL_IMAGE:figures/full_fig_p114_28.png] view at source ↗
Figure 29
Figure 29. Figure 29: FIG. 29: SIR dynamics and the associated eigenvalue spectra, with and without a network hub. [PITH_FULL_IMAGE:figures/full_fig_p122_29.png] view at source ↗
Figure 30
Figure 30. Figure 30: FIG. 30: Eigenvalue density of the Laplacian matrix with [PITH_FULL_IMAGE:figures/full_fig_p124_30.png] view at source ↗
Figure 31
Figure 31. Figure 31: FIG. 31: Turing instability on the 1D chain and the ER graph with [PITH_FULL_IMAGE:figures/full_fig_p127_31.png] view at source ↗
Figure 32
Figure 32. Figure 32: FIG. 32: Eigenvalue density of the scaled adjacency matrix of a configuration model network. [PITH_FULL_IMAGE:figures/full_fig_p129_32.png] view at source ↗
Figure 33
Figure 33. Figure 33: FIG. 33: Eigenvalue density of the sum of a GOE matrix and a standard Wischart matrix. The [PITH_FULL_IMAGE:figures/full_fig_p145_33.png] view at source ↗
Figure 34
Figure 34. Figure 34: FIG. 34: Eigenvalue density of the product of a standard Wischart and a GOE matrix. The [PITH_FULL_IMAGE:figures/full_fig_p147_34.png] view at source ↗
Figure 35
Figure 35. Figure 35: FIG. 35: (Top) The diagrammatic representation of the quartic term before pairing. (Bottom) [PITH_FULL_IMAGE:figures/full_fig_p156_35.png] view at source ↗
Figure 36
Figure 36. Figure 36: FIG. 36: Possible planar Wick pairings for the second term in the expansion of [PITH_FULL_IMAGE:figures/full_fig_p157_36.png] view at source ↗
Figure 37
Figure 37. Figure 37: FIG. 37: (Left) The order parameters [PITH_FULL_IMAGE:figures/full_fig_p166_37.png] view at source ↗
Figure 38
Figure 38. Figure 38: FIG. 38: Critical temperature as a function of [PITH_FULL_IMAGE:figures/full_fig_p167_38.png] view at source ↗
Figure 39
Figure 39. Figure 39: FIG. 39: Eigenvalue density of the GUE ensemble, with [PITH_FULL_IMAGE:figures/full_fig_p185_39.png] view at source ↗
Figure 40
Figure 40. Figure 40: FIG. 40: The sum over all possible diagrams. Recognising the self-similarity of the series, this can [PITH_FULL_IMAGE:figures/full_fig_p207_40.png] view at source ↗
Figure 41
Figure 41. Figure 41: FIG. 41: Eigenvalue density of the scaled adjacency matrices (left) with [PITH_FULL_IMAGE:figures/full_fig_p212_41.png] view at source ↗

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

274 extracted references · 274 canonical work pages · 9 internal anchors

  1. [1]

    It is also possible to make a perturbative expansion of the response functions in model parameters other than 1/Nusing the diagrammatic approach [56, 187] (see also the exercises)

    using the Keldysh dynamic formalism [248, 249]). It is also possible to make a perturbative expansion of the response functions in model parameters other than 1/Nusing the diagrammatic approach [56, 187] (see also the exercises). Here, we use the computation of the semicircle law to demonstrate the diagrammatic approach. Needless to say, the diagrammatic ...

    work page

  2. [2]

    Evaluation using Wick’s theorem We begin with Eq. (547). Writing the path integral explicitly, and consulting Eq. (549), we may write the disorder-averaged response functions (in the caseµ= 0 and Γ = 1) as * 1 N X k Rkk(T,0) + = Z D[x,ˆx] − i N X k xk(T)ˆxk(0) ! exp [S0 +S int],(572) where we identify the so-called ‘bare’ action and the interaction term r...

    work page

  3. [3]

    (574) is a daunting task

    Feynman diagrams as a combinatorial tool Keeping track of the huge variety of ‘Wick pairings’ in the sum in Eq. (574) is a daunting task. A useful strategy is to use Feynman diagrams, which help to identify the terms that are (non- )vanishing in the limitN→ ∞, as well as being a convenient bookkeeping tool. We have already seen that Wick pairings can vani...

    work page

  4. [4]

    The only Wick pairings that we need to consider pair solely hatted and unhatted dynamic variables

    work page

  5. [5]

    The only non-vanishing Wick pairings forN→ ∞correspond to planar diagrams with non-crossing and non-twisted arcs

    work page

  6. [6]

    One therefore sees that the sum in Eq

    The number of combinations of Wick pairings that are equivalent up to time ordering always exactly cancels a prefactor, allowing us to discard the labelling of the internal nodes in the Feynman diagrams. One therefore sees that the sum in Eq. (574) can be evaluated in the thermodynamic limit by considering the set of all planar rainbow diagrams. As a fina...

    work page

  7. [7]

    i X i Z dtψixi − σ2 T 2 X i Z dtˆx2 i (t) # ×exp

    Resummation of the diagrammatic series We employ one additional diagrammatic convention to simplify the notation when we perform sums over many diagrams. We denote a sum of planar diagrams by an edge with a double arrow, accompanied by a label for identification purposes. For example, let us take the surviving planar diagrams for the second-order term abo...

    work page

  8. [8]

    Initialise a populationG (cav) i with random complex values andi= 1,· · ·, N pop

    work page

  9. [9]

    For eachi, draw a value ofk i from the distributionkP k/p

    work page

  10. [10]

    For eachi, select at random a subsetS i of sizek i −1 from the rangej= 1,· · ·, N pop, and calculateG ′ i = 1 ω−iϵ− 1 p P j∈Si G(cav) j

    work page

  11. [11]

    Repeat from Step 2 until the quantityG cav =N −1 pop P i G(cav) i converges

    work page

  12. [12]

    Perform Steps 2 to 4 once more

    Once convergence has occurred, we begin the sampling phase. Perform Steps 2 to 4 once more. 213

    work page

  13. [13]

    Select a random subsetS j of sizek j from the rangel= 1,· · ·, N pop

    Draw a valuek j from the distributionP k. Select a random subsetS j of sizek j from the rangel= 1,· · ·, N pop. ComputeG j = 1 ω−iη− 1 p P l∈Sj G(cav) l

    work page

  14. [14]

    Compute the quantityG α = N −1 samp PNsamp j=1 Gj

    Repeat Step 7 a numberN samp times to obtain a set{G j}. Compute the quantityG α = N −1 samp PNsamp j=1 Gj

    work page

  15. [15]

    ComputeG(ω) = N −1 iter PNiter α=1 Gα

    Repeat Steps 6 to 8 a numberN iter times to obtain a set{G α}. ComputeG(ω) = N −1 iter PNiter α=1 Gα. This is our final estimate of the resolvent for a fixed valueω

    work page

  16. [16]

    The eigenvalue density is then obtained via ρ(ω) =π −1ImG(ω)

    Repeat Steps 1 to 9 for all desired values ofω. The eigenvalue density is then obtained via ρ(ω) =π −1ImG(ω). The results of performing this procedure are show in Fig. 41. To adapt the method for the Laplacian matrix, we instead use the update ruleG ′ i = 1 ω−iϵ+ 1 p P j∈Si 1 1+G(cav) j /p . Let us discuss briefly the rationale behind this algorithm. For ...

    work page

  17. [17]

    M. L. Mehta,Random matrices, Vol. 142 (Elsevier, 2004)

    work page 2004

  18. [18]

    R. Haq, A. Pandey, and O. Bohigas, Fluctuation properties of nuclear energy levels: Do theory and experiment agree?, Physical Review Letters48, 1086 (1982)

    work page 1982

  19. [19]

    Weidenm¨ uller and G

    H. Weidenm¨ uller and G. Mitchell, Random matrices and chaos in nuclear physics: Nuclear structure, Reviews of Modern Physics81, 539 (2009)

    work page 2009

  20. [20]

    R. M. May, Will a large complex system be stable?, Nature238, 413 (1972). 214

    work page 1972

  21. [21]

    Allesina and S

    S. Allesina and S. Tang, The stability–complexity relationship at age 40: a random matrix perspective, Population Ecology57, 63 (2015)

    work page 2015

  22. [22]

    J. Baik, G. Ben Arous, and S. P´ ech´ e, Phase transition of the largest eigenvalue for nonnull complex sample covariance matrices, Ann. Probab.33, 1643 (2005)

    work page 2005

  23. [23]

    Laloux, P

    L. Laloux, P. Cizeau, M. Potters, and J.-P. Bouchaud, Random matrix theory and financial correla- tions, International Journal of Theoretical and Applied Finance3, 391 (2000)

    work page 2000

  24. [24]

    Financial Applications of Random Matrix Theory: Old Laces and New Pieces

    M. Potters, J.-P. Bouchaud, and L. Laloux, Financial applications of random matrix theory: Old laces and new pieces, arXiv preprint physics/0507111 (2005)

    work page internal anchor Pith review Pith/arXiv arXiv 2005

  25. [25]

    M´ ezard, G

    M. M´ ezard, G. Parisi, and M. A. Virasoro,Spin glass theory and beyond: An Introduction to the Replica Method and Its Applications, Vol. 9 (World Scientific Publishing Company, 1987)

    work page 1987

  26. [26]

    Evers and A

    F. Evers and A. D. Mirlin, Anderson transitions, Reviews of Modern Physics80, 1355 (2008)

    work page 2008

  27. [27]

    Bohigaset al.,Random matrix theories and chaotic dynamics, Tech

    O. Bohigaset al.,Random matrix theories and chaotic dynamics, Tech. Rep. (Paris-11 Univ., 91-Orsay (France). Inst. de Physique Nucleaire, 1991)

    work page 1991

  28. [28]

    Bohigas and M.-J

    O. Bohigas and M.-J. Giannoni, Chaotic motion and random matrix theories, inMathematical and Computational Methods in Nuclear Physics: Proceedings of the Sixth Granada Workshop Held in Granada, Spain, October 3–8, 1983(Springer, 2005) pp. 1–99

    work page 1983

  29. [29]

    Di Francesco, P

    P. Di Francesco, P. Ginsparg, and J. Zinn-Justin, 2d gravity and random matrices, Physics Reports 254, 1 (1995)

    work page 1995

  30. [30]

    J. P. Keating and N. C. Snaith, Random matrix theory andζ(1/2+ it), Communications in Mathe- matical Physics214, 57 (2000)

    work page 2000

  31. [31]

    A. M. Odlyzko, On the distribution of spacings between zeros of the zeta function, Mathematics of Computation48, 273 (1987)

    work page 1987

  32. [32]

    Johansson, Non-intersecting paths, random tilings and random matrices, Probability theory and related fields123, 225 (2002)

    K. Johansson, Non-intersecting paths, random tilings and random matrices, Probability theory and related fields123, 225 (2002)

    work page 2002

  33. [33]

    E. P. Wigner, Characteristic vectors of bordered matrices with infinite dimensions, Annals of Mathe- matics62, 548 (1955)

    work page 1955

  34. [34]

    Wigner, Conference on neutron physics by time-of-flight, Oak Ridge National Lab

    E. Wigner, Conference on neutron physics by time-of-flight, Oak Ridge National Lab. Report ORNL- 2309 (1957)

    work page 1957

  35. [35]

    E. P. Wigner, On the distribution of the roots of certain symmetric matrices, Annals of Mathematics 67, 325 (1958)

    work page 1958

  36. [36]

    E. P. Wigner, Random matrices in physics, SIAM review9, 1 (1967)

    work page 1967

  37. [37]

    E. P. Wigner, Characteristic vectors of bordered matrices with infinite dimensions i, inThe Collected Works of Eugene Paul Wigner: Part A: The Scientific Papers(Springer, 1993) pp. 524–540

    work page 1993

  38. [38]

    D. V. Widder, The stieltjes transform, Transactions of the American Mathematical Society43, 7 (1938)

    work page 1938

  39. [39]

    K¨ uhn, Spectra of sparse random matrices, Journal of Physics A: Mathematical and Theoretical41, 295002 (2008)

    R. K¨ uhn, Spectra of sparse random matrices, Journal of Physics A: Mathematical and Theoretical41, 295002 (2008). 215

    work page 2008

  40. [40]

    Rogers, I

    T. Rogers, I. P. Castillo, R. K¨ uhn, and K. Takeda, Cavity approach to the spectral density of sparse symmetric random matrices, Physical Review E—Statistical, Nonlinear, and Soft Matter Physics78, 031116 (2008)

    work page 2008

  41. [41]

    Gaudin, Sur la loi limite de l’espacement des valeurs propres d’une matrice ale´ atoire, Nuclear Physics25, 447 (1961)

    M. Gaudin, Sur la loi limite de l’espacement des valeurs propres d’une matrice ale´ atoire, Nuclear Physics25, 447 (1961)

    work page 1961

  42. [42]

    Efetov, Supersymmetry and theory of disordered metals, advances in Physics32, 53 (1983)

    K. Efetov, Supersymmetry and theory of disordered metals, advances in Physics32, 53 (1983)

    work page 1983

  43. [43]

    Erd˝ os, Universality of wigner random matrices: a survey of recent results, Russian Mathematical Surveys66, 507 (2011)

    L. Erd˝ os, Universality of wigner random matrices: a survey of recent results, Russian Mathematical Surveys66, 507 (2011)

    work page 2011

  44. [44]

    A. D. Mirlin and Y. V. Fyodorov, Universality of level correlation function of sparse random matrices, Journal of Physics A: Mathematical and General24, 2273 (1991)

    work page 1991

  45. [45]

    F. J. Dyson, Statistical theory of the energy levels of complex systems. i, Journal of Mathematical Physics3, 140 (1962)

    work page 1962

  46. [46]

    F. J. Dyson, The threefold way. algebraic structure of symmetry groups and ensembles in quantum mechanics, Journal of Mathematical Physics3, 1199 (1962)

    work page 1962

  47. [47]

    F. J. Dyson, Statistical theory of the energy levels of complex systems. iii, Journal of Mathematical Physics3, 166 (1962)

    work page 1962

  48. [48]

    Bohigas, M.-J

    O. Bohigas, M.-J. Giannoni, and C. Schmit, Characterization of chaotic quantum spectra and univer- sality of level fluctuation laws, Physical review letters52, 1 (1984)

    work page 1984

  49. [49]

    H. L. Montgomery, The pair correlation of zeros of the zeta function, inProc. Symp. Pure Math, Vol. 24 (1973) p. 1

    work page 1973

  50. [50]

    Rajan and L

    K. Rajan and L. F. Abbott, Eigenvalue spectra of random matrices for neural networks, Physical review letters97, 188104 (2006)

    work page 2006

  51. [51]

    Sompolinsky, A

    H. Sompolinsky, A. Crisanti, and H.-J. Sommers, Chaos in random neural networks, Physical review letters61, 259 (1988)

    work page 1988

  52. [52]

    Aljadeff, M

    J. Aljadeff, M. Stern, and T. Sharpee, Transition to chaos in random networks with cell-type-specific connectivity, Physical review letters114, 088101 (2015)

    work page 2015

  53. [53]

    Molgedey, J

    L. Molgedey, J. Schuchhardt, and H. G. Schuster, Suppressing chaos in neural networks by noise, Physical review letters69, 3717 (1992)

    work page 1992

  54. [54]

    Haake, F

    F. Haake, F. Izrailev, N. Lehmann, D. Saher, and H.-J. Sommers, Statistics of complex levels of random matrices for decaying systems, Zeitschrift f¨ ur Physik B Condensed Matter88, 359 (1992)

    work page 1992

  55. [55]

    Feinberg and A

    J. Feinberg and A. Zee, Non-hermitian random matrix theory: Method of hermitian reduction, Nuclear Physics B504, 579 (1997)

    work page 1997

  56. [56]

    R. M. May,Stability and complexity in model ecosystems, Vol. 6 (Princeton university press, 2001)

    work page 2001

  57. [57]

    M. A. Nowak and W. Tarnowski, Probing non-orthogonality of eigenvectors in non-hermitian matrix models: diagrammatic approach, Journal of High Energy Physics2018, 1 (2018)

    work page 2018

  58. [58]

    V. L. Girko, Circular law, Theory of Probability & Its Applications29, 694 (1985)

    work page 1985

  59. [59]

    Girko, Elliptic law, Theory of Probability & Its Applications30, 677 (1986)

    V. Girko, Elliptic law, Theory of Probability & Its Applications30, 677 (1986). 216

    work page 1986

  60. [60]

    R. A. Janik, M. A. Nowak, G. Papp, and I. Zahed, Non-hermitian random matrix models, Nuclear Physics B501, 603 (1997)

    work page 1997

  61. [61]

    H. J. Sommers, A. Crisanti, H. Sompolinsky, and Y. Stein, Spectrum of large random asymmetric matrices, Physical review letters60, 1895 (1988)

    work page 1988

  62. [62]

    Rogers, Universal sum and product rules for random matrices, Journal of mathematical physics51 (2010)

    T. Rogers, Universal sum and product rules for random matrices, Journal of mathematical physics51 (2010)

    work page 2010

  63. [63]

    Benaych-Georges and R

    F. Benaych-Georges and R. R. Nadakuditi, The eigenvalues and eigenvectors of finite, low rank per- turbations of large random matrices, Advances in Mathematics227, 494 (2011)

    work page 2011

  64. [64]

    O’Rourke and D

    S. O’Rourke and D. Renfrew, Low rank perturbations of large elliptic random matrices, (2014)

    work page 2014

  65. [65]

    J. W. Baron, T. J. Jewell, C. Ryder, and T. Galla, Eigenvalues of random matrices with generalized correlations: A path integral approach, Physical Review Letters128, 120601 (2022)

    work page 2022

  66. [66]

    J. W. Baron, T. J. Jewell, C. Ryder, and T. Galla, Breakdown of random-matrix universality in persistent lotka-volterra communities, Physical Review Letters130, 137401 (2023)

    work page 2023

  67. [67]

    S. F. Edwards and R. C. Jones, The eigenvalue spectrum of a large symmetric random matrix, Journal of Physics A: Mathematical and General9, 1595 (1976)

    work page 1976

  68. [68]

    Ikeda, Bose–einstein-like condensation of deformed random matrix: a replica approach, Journal of Statistical Mechanics: Theory and Experiment2023, 023302 (2023)

    H. Ikeda, Bose–einstein-like condensation of deformed random matrix: a replica approach, Journal of Statistical Mechanics: Theory and Experiment2023, 023302 (2023)

    work page 2023

  69. [69]

    J. Hu, D. R. Amor, M. Barbier, G. Bunin, and J. Gore, Emergent phases of ecological diversity and dynamics mapped in microcosms, Science378, 85 (2022)

    work page 2022

  70. [70]

    G. J. Rodgers, K. Austin, B. Kahng, and D. Kim, Eigenvalue spectra of complex networks, Journal of Physics A: Mathematical and General38, 9431 (2005)

    work page 2005

  71. [71]

    J. W. Baron, Eigenvalue spectra and stability of directed complex networks, Physical Review E106, 064302 (2022)

    work page 2022

  72. [72]

    J. W. Baron, Path-integral approach to sparse non-hermitian random matrices, Physical Review E 111, 034217 (2025)

    work page 2025

  73. [73]

    Valigi, J

    P. Valigi, J. W. Baron, I. Neri, G. Biroli, and C. Cammarota, Eigenvalue spectral tails and localisation properties of asymmetric networks, Journal of Physics A: Mathematical and Theoretical58, 455002 (2025)

    work page 2025

  74. [74]

    J. W. Baron and T. Galla, Dispersal-induced instability in complex ecosystems, Nature communica- tions11, 6032 (2020)

    work page 2020

  75. [75]

    Allesina, J

    S. Allesina, J. Grilli, G. Barab´ as, S. Tang, J. Aljadeff, and A. Maritan, Predicting the stability of large structured food webs, Nature communications6, 7842 (2015)

    work page 2015

  76. [76]

    Poley, T

    L. Poley, T. Galla, and J. W. Baron, Eigenvalue spectra of finely structured random matrices, Physical Review E109, 064301 (2024)

    work page 2024

  77. [77]

    Barab´ as, M

    G. Barab´ as, M. J. Michalska-Smith, and S. Allesina, Self-regulation and the stability of large ecological networks, Nature ecology & evolution1, 1870 (2017)

    work page 2017

  78. [78]

    Pigani, D

    E. Pigani, D. Sgarbossa, S. Suweis, A. Maritan, and S. Azaele, Delay effects on the stability of large 217 ecosystems, Proceedings of the National Academy of Sciences119, e2211449119 (2022)

    work page 2022

  79. [79]

    Gravel, F

    D. Gravel, F. Massol, and M. A. Leibold, Stability and complexity in model meta-ecosystems, Nature communications7, 12457 (2016)

    work page 2016

  80. [80]

    Interaction patterns and diversity in assembled ecological communities

    G. Bunin, Interaction patterns and diversity in assembled ecological communities, arXiv preprint arXiv:1607.04734 (2016)

    work page internal anchor Pith review Pith/arXiv arXiv 2016

Showing first 80 references.