pith. sign in

arxiv: 2510.22120 · v2 · submitted 2025-10-25 · 🧮 math-ph · hep-th· math.MP· math.PR

A Two-HCIZ Gaussian Matrix Model for Non-intersecting Brownian Bridges

Maksim Kosmakov This is my paper

Pith reviewed 2026-05-18 05:19 UTC · model grok-4.3

classification 🧮 math-ph hep-thmath.MPmath.PR
keywords non-intersecting Brownian bridgesHCIZ integralKarlin-McGregor lawHermitian matrix ensemblesmultiple orthogonal polynomialsRiemann-Hilbert problemSchwinger-Dyson equations
0
0 comments X

The pith

A unitarily invariant Hermitian matrix ensemble is built from two HCIZ integrals so that its eigenvalues at fixed time follow the Karlin-McGregor law for non-intersecting Brownian bridges with any finite endpoint multiplicities.

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

The paper constructs a random matrix model whose eigenvalue distribution exactly reproduces the known determinant formula for non-intersecting Brownian bridges that start and end in clusters of arbitrary sizes. This supplies an explicit Hermitian ensemble realization for a problem previously handled through mixed-type multiple orthogonal polynomials and associated Riemann-Hilbert problems. The construction yields several exact finite-n identities, including a path-space lift to an orbital Hermitian Brownian bridge and a reduction of the partition function to one compact HCIZ integral whose time dependence is explicit.

Core claim

The authors define a Hermitian matrix ensemble whose weight is given by a specific linear combination of two Gaussian-weighted HCIZ integrals. At any fixed time the joint eigenvalue density of this ensemble coincides with the Karlin-McGregor determinant that governs non-intersecting Brownian bridges with prescribed finite multiplicities at the initial and terminal times. The same weight also produces a reduction of the partition function to a single HCIZ integral and admits a lift to a matrix-valued Brownian bridge whose eigenvalues never intersect.

What carries the argument

A linear combination of two Harish-Chandra-Itzykson-Zuber (HCIZ) integrals with Gaussian weights, chosen so that the resulting eigenvalue marginal equals the Karlin-McGregor determinant for arbitrary finite endpoint multiplicities.

If this is right

  • The ensemble admits an orbital Hermitian Brownian bridge interpretation on the full path space.
  • The partition function reduces to a single compact HCIZ integral with explicit t-dependence.
  • The one-sided limit is spectrally equivalent to the Gaussian external-field ensemble yet possesses different angular statistics.
  • Fixed-time Schwinger-Dyson identities and associated resolvent relations hold for the dressed ensemble.

Where Pith is reading between the lines

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

  • The angular mismatch between the two spectrally equivalent ensembles suggests that full-matrix correlation functions beyond eigenvalues will differ and could be computed from the HCIZ representation.
  • The explicit HCIZ form may allow transfer of asymptotic techniques from integrable systems to obtain large-n limits for the multi-start/multi-end bridge problem without solving the Riemann-Hilbert problem anew.
  • The construction supplies a concrete random-matrix dynamics whose eigenvalue non-intersection is enforced by the unitary invariance and the choice of weights, offering a possible route to study the full process as a matrix-valued diffusion.

Load-bearing premise

There exists a specific combination of two HCIZ integrals with Gaussian weights whose induced eigenvalue law exactly matches the Karlin-McGregor determinant for every choice of finite multiplicities at both endpoints.

What would settle it

Direct numerical evaluation, for small matrix size n and chosen endpoint multiplicities, of the eigenvalue density produced by the two-HCIZ weight and comparison against the explicit Karlin-McGregor determinant or the zeros of the corresponding multiple orthogonal polynomials.

read the original abstract

We construct a unitarily invariant Hermitian matrix ensemble whose fixed-time eigenvalue law coincides with the Karlin--McGregor law for non-intersecting Brownian bridges with arbitrary finite multiplicities at both endpoints. This provides an explicit matrix-ensemble realization of the known mixed-type multiple orthogonal polynomial and Riemann--Hilbert description of the general multi-start/multi-end problem. We then derive several exact finite-$n$ consequences of this construction. These include a path-space lift as an orbital Hermitian Brownian bridge and a reduction of the partition function to a single compact HCIZ integral with explicit $t$-dependence. We also compare the one-sided reduction with the Gaussian external-field ensemble, showing that, although the two ensembles are spectrally equivalent, their angular statistics are different. Finally, we derive fixed-time Schwinger--Dyson identities and associated resolvent relations for the dressed ensemble.

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 constructs a unitarily invariant Hermitian matrix ensemble whose fixed-time eigenvalue law is asserted to coincide with the Karlin-McGregor determinant formula for non-intersecting Brownian bridges with arbitrary finite multiplicities at both endpoints. This is achieved via a specific combination of two HCIZ integrals with Gaussian weights, providing an explicit matrix-ensemble realization of the mixed-type multiple orthogonal polynomial and Riemann-Hilbert description. The paper then derives exact finite-n consequences, including a path-space lift as an orbital Hermitian Brownian bridge, a reduction of the partition function to a single compact HCIZ integral with explicit t-dependence, a comparison of the one-sided reduction with the Gaussian external-field ensemble (spectrally equivalent but with different angular statistics), and fixed-time Schwinger-Dyson identities with associated resolvent relations.

Significance. If the central identification holds rigorously, the work supplies a concrete random-matrix realization for the general multi-start/multi-end non-intersecting bridge problem, linking Hermitian matrix models to multiple orthogonal polynomials and integrable systems. The exact finite-n results, the single-HCIZ reduction, and the Schwinger-Dyson identities constitute useful additions that could enable further exact calculations. The distinction drawn between spectral equivalence and differing angular statistics in the one-sided case is a clear and worthwhile observation.

major comments (2)
  1. [§2] §2 (central construction): The claim that a specific combination of two HCIZ integrals with Gaussian weights yields, after unitary integration, an eigenvalue marginal whose density is precisely the Karlin-McGregor determinant det(p(t; a_i, x_j)) det(p(T-t; x_j, b_k)) (or its multiple-orthogonal generalization) times the Vandermonde factor for arbitrary finite endpoint multiplicities requires an explicit weight function and a direct term-by-term or limiting comparison against the coalesced-point formula. The current presentation leaves this identification at the level of an existence statement; without this verification the central claim is not yet load-bearing.
  2. [§4] §4 (partition-function reduction): The reduction to a single compact HCIZ integral with explicit t-dependence is presented as a consequence of the two-HCIZ construction; if the multiplicity handling in §2 relies on an unstated regularization or limit, the reduction formula and its t-dependence must be re-derived or qualified to remain valid for coalesced endpoints.
minor comments (2)
  1. [Notation] Notation for the two HCIZ integrals and the precise Gaussian weights should be introduced with explicit integral expressions at the first appearance to improve readability.
  2. [Introduction] A short remark on how the construction recovers the classical distinct-endpoint Karlin-McGregor case as a special instance would help orient readers.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the careful reading of the manuscript and the constructive comments. We respond to each major comment below.

read point-by-point responses

  1. Referee: [§2] §2 (central construction): The claim that a specific combination of two HCIZ integrals with Gaussian weights yields, after unitary integration, an eigenvalue marginal whose density is precisely the Karlin-McGregor determinant det(p(t; a_i, x_j)) det(p(T-t; x_j, b_k)) (or its multiple-orthogonal generalization) times the Vandermonde factor for arbitrary finite endpoint multiplicities requires an explicit weight function and a direct term-by-term or limiting comparison against the coalesced-point formula. The current presentation leaves this identification at the level of an existence statement; without this verification the central claim is not yet load-bearing.

    Authors: We agree that an explicit verification strengthens the central claim. The ensemble is defined by the indicated product of HCIZ integrals; after the unitary integration the eigenvalue weight is obtained by the standard Harish-Chandra formula applied to each factor. In the revised manuscript we will write the resulting explicit weight function on the eigenvalues and carry out the direct (limiting) comparison with the coalesced Karlin–McGregor determinant, thereby making the identification fully rigorous rather than existential. revision: yes

  2. Referee: [§4] §4 (partition-function reduction): The reduction to a single compact HCIZ integral with explicit t-dependence is presented as a consequence of the two-HCIZ construction; if the multiplicity handling in §2 relies on an unstated regularization or limit, the reduction formula and its t-dependence must be re-derived or qualified to remain valid for coalesced endpoints.

    Authors: The reduction in §4 is obtained by performing the two unitary integrals in sequence and collecting the resulting Gaussian factors; the explicit t-dependence follows from the quadratic terms in the exponents. In the revision we will state the limiting procedure for finite multiplicities already at the level of the two-HCIZ definition and re-derive the single-HCIZ formula under that limit, confirming that the t-dependence survives unchanged. This qualification will ensure validity for coalesced endpoints. revision: yes

Circularity Check

0 steps flagged

No significant circularity: construction matches external Karlin-McGregor law via explicit HCIZ combination without self-referential fitting or load-bearing self-citation.

full rationale

The paper defines a specific two-HCIZ Gaussian matrix ensemble and derives that its fixed-time eigenvalue marginal equals the known Karlin-McGregor determinant formula (including multiplicity handling) as a consequence of the unitary invariance and HCIZ integral properties. This is an external matching to a pre-existing probabilistic law rather than a self-definition or fitted parameter renamed as prediction. No self-citation chain is invoked to justify the central identification, and the derivation remains self-contained against the independent Karlin-McGregor reference. The abstract and construction sketch present the coincidence as a derived result, not an input assumption.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

The central claim rests on the existence of a unitarily invariant ensemble whose marginal matches a known determinant formula; no free parameters, axioms, or invented entities are visible in the abstract.

pith-pipeline@v0.9.0 · 5677 in / 1126 out tokens · 25888 ms · 2026-05-18T05:19:04.408584+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

25 extracted references · 25 canonical work pages · 1 internal anchor

  1. [1]

    Hermitian, symmetric and symplectic random ensembles: PDEs for the distribution of the spectrum

    Mark Adler and Pierre van Moerbeke. “Hermitian, symmetric and symplectic random ensembles: PDEs for the distribution of the spectrum”.Annals of Mathematics153.1 (2001), pp. 149–189

    work page 2001

  2. [2]

    Non-intersecting Brownian motionsleavingfromandgoingtoseveralpoints

    Mark Adler, Pierre van Moerbeke, and Didier Vanderstichelen. “Non-intersecting Brownian motionsleavingfromandgoingtoseveralpoints”.PhysicaD:NonlinearPhenomena241.5(2012), pp. 443–460

    work page 2012

  3. [3]

    Random matrices, Virasoro algebras, and noncommutative KP

    Mark Adler, Takahiro Shiota, and Pierre van Moerbeke. “Random matrices, Virasoro algebras, and noncommutative KP”.Duke Mathematical Journal94.2 (Aug. 1998), pp. 379–431

    work page 1998

  4. [4]

    Note sur une relation entre les intégrales définies des produits des fonctions

    C. Andréief. “Note sur une relation entre les intégrales définies des produits des fonctions”. French. Mém. Soc. Sci. Phys. Nat. Bordeaux (Sér. 3)2 (1886), pp. 1–14

    work page

  5. [5]

    Another Christoffel–Darboux formula for multiple orthogonal polynomials of mixed type

    Gonzalo Ariznabarreta and Manuel Mañas. “Another Christoffel–Darboux formula for multiple orthogonal polynomials of mixed type”.Linear Algebra and its Applications468 (2015), pp. 154– 170

    work page 2015

  6. [6]

    Random-matrix theory of quantum transport

    C. W. J. Beenakker. “Random-matrix theory of quantum transport”.Reviews of Modern Physics 69.3 (1997), pp. 731–808

    work page 1997

  7. [7]

    Planar orthogonal polynomials as type I multiple orthogonal polynomials

    Sergey Berezin, Arno B. J. Kuijlaars, and Isabel Parra. “Planar orthogonal polynomials as type I multiple orthogonal polynomials”.SIGMA19 (2023), Paper 020. 20

    work page 2023

  8. [8]

    Random matrix model with external source and a constrained vector equilibrium problem

    Pavel Bleher, Steven Delvaux, and Arno B. J. Kuijlaars. “Random matrix model with external source and a constrained vector equilibrium problem”.Communications on Pure and Applied Mathematics64.1 (2011), pp. 116–160

    work page 2011

  9. [9]

    Lectures on random matrix models. The Riemann-Hilbert approach

    Pavel M. Bleher.Lectures on random matrix models. The Riemann–Hilbert approach. Chapter in RandomMatrices,RandomProcessesandIntegrableSystems,CRMSeriesinMathematicalPhysics, Springer. 2011. eprint:\href{https://arxiv.org/abs/0801.1858}{arXiv:0801.1858}

    work page internal anchor Pith review Pith/arXiv arXiv 2011

  10. [10]

    Biorthogonal ensembles

    Alexei Borodin. “Biorthogonal ensembles”.Nuclear Physics B536.3 (1998), pp. 704–732

    work page 1998

  11. [11]

    Multiple orthogonal polynomials of mixed type and non-intersecting Brownian motions

    Ettore Daems and Arno B. J. Kuijlaars. “Multiple orthogonal polynomials of mixed type and non-intersecting Brownian motions”.Journal of Approximation Theory146.1 (2007), pp. 91–114

    work page 2007

  12. [12]

    A steepest descent method for oscillatory Riemann–Hilbert problems. Asymptotics for the MKdV equation

    Percy Deift and Xin Zhou. “A steepest descent method for oscillatory Riemann–Hilbert problems. Asymptotics for the MKdV equation”.Annals of Mathematics137.2 (1993), pp. 295–368

    work page 1993

  13. [13]

    A Brownian-motion model for the eigenvalues of a random matrix

    Freeman J. Dyson. “A Brownian-motion model for the eigenvalues of a random matrix”.Journal of Mathematical Physics3.6 (1962), pp. 1191–1198

    work page 1962

  14. [14]

    Differential operators on a semisimple Lie algebra

    Harish-Chandra. “Differential operators on a semisimple Lie algebra”.American Journal of Mathematics79.1 (1957), pp. 87–120

    work page 1957

  15. [15]

    Scalarproductsofsymmetricfunctionsandmatrixintegrals

    JohnHarnadandAlexanderY.Orlov.“Scalarproductsofsymmetricfunctionsandmatrixintegrals”. Theoretical and Mathematical Physics137.3 (2003), pp. 1676–1690

    work page 2003

  16. [16]

    Matrix integration and combinatorics of modular groups

    Claude Itzykson and Jean-Bernard Zuber. “Matrix integration and combinatorics of modular groups”.Communications in Mathematical Physics134.1 (1990), pp. 197–207

    work page 1990

  17. [17]

    Theplanarapproximation.II

    ClaudeItzyksonandJean-BernardZuber.“Theplanarapproximation.II”.JournalofMathematical Physics21.3 (1980), pp. 411–421

    work page 1980

  18. [18]

    Matrix models for eigenstate thermalization

    Dionysios L. Jafferis et al. “Matrix models for eigenstate thermalization”.Physical Review X13.3 (2023), p. 031033

    work page 2023

  19. [19]

    Coincidence probabilities

    Samuel Karlin and James McGregor. “Coincidence probabilities”.Pacific Journal of Mathematics 9.4 (1959), pp. 1141–1164

    work page 1959

  20. [20]

    Multiple orthogonal polynomials in random matrix theory

    Arno B. J. Kuijlaars. “Multiple orthogonal polynomials in random matrix theory”.Proceedings of the International Congress of Mathematicians. Vol. III. Hyderabad ICM 2010. 2010, pp. 1417– 1432

    work page 2010

  21. [21]

    Madan Lal Mehta.Random Matrices. 3rd. Vol. 142. Pure and Applied Mathematics. Amsterdam: Elsevier/Academic Press, 2004.isbn: 978-0-12-088409-4

    work page 2004

  22. [22]

    Tau functions and matrix integrals

    Alexander Yu. Orlov. “Tau functions and matrix integrals”.Theoretical and Mathematical Physics 128.1 (2001), pp. 906–926

    work page 2001

  23. [23]

    Mathematical Surveys and Monographs

    LeonidPasturandMariyaShcherbina.EigenvalueDistributionofLargeRandomMatrices.Vol.171. Mathematical Surveys and Monographs. American Mathematical Society, 2011

    work page 2011

  24. [24]

    HCIZ integral and 2D Toda lattice hierarchy

    P. Zinn-Justin. “HCIZ integral and 2D Toda lattice hierarchy”.Nuclear Physics B634.3 (2002), pp. 417–432

    work page 2002

  25. [25]

    Random Hermitian matrices in an external field

    P. Zinn-Justin. “Random Hermitian matrices in an external field”.Nuclear Physics B497.3 (1997), pp. 725–732. 21

    work page 1997