Thermal Concentration and Poisson--Dirichlet Edge Statistics for Random--Lattice Gibbs Ensembles

Masahiro Kaminaga

arxiv: 2607.00311 · v1 · pith:S5QEJLIQnew · submitted 2026-07-01 · 🧮 math.PR · math-ph· math.MP

Thermal Concentration and Poisson--Dirichlet Edge Statistics for Random--Lattice Gibbs Ensembles

Masahiro Kaminaga This is my paper

Pith reviewed 2026-07-02 00:57 UTC · model grok-4.3

classification 🧮 math.PR math-phmath.MP MSC 60G5582B0511H06

keywords random latticesGibbs measuresPoisson-Dirichlet distributionshortest vector problemthermal concentrationHaar measureunimodular lattices

0 comments

The pith

Gibbs ensembles on random lattices yield Poisson-Dirichlet edge statistics above inverse temperature 1 and thermal concentration for primitive directions at c equals gamma to the minus 2.

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

The paper establishes limit theorems for Gibbs measures on high-dimensional Haar-random unimodular lattices, where energy is the squared Euclidean norm of lattice vectors and c scales the inverse temperature. For the full sign-class ensemble, the Gibbs mass in an edge window of near-shortest vectors vanishes in probability for c at most 1, converges to a Poisson process for c greater than 1, and the ranked weights follow the Poisson-Dirichlet law with parameter 1 over c. For a primitive-direction ensemble with fixed approximation factor gamma greater than 1, a weighted moment formula and quenched concentration hold in the high-temperature regime, producing a visibility threshold at c equals gamma to the minus 2 where the primitive mass jumps from zero to one.

Core claim

For the full sign-class Gibbs ensemble the edge-window mass vanishes in probability when 0 less than c less than or equal to 1, admits a nontrivial Poisson limit when c greater than 1, and the ordered Gibbs weights converge to the Poisson-Dirichlet distribution with parameter 1 over c; separately, the primitive fixed-factor Gibbs mass tends to zero for c less than gamma to the minus 2, to one for gamma to the minus 2 less than c less than 1, and equals one half at the critical value c equals gamma to the minus 2.

What carries the argument

The Gibbs measure on quenched Haar-random unimodular lattices with energy given by squared Euclidean length, restricted to an edge window scaled by e to the a over n.

If this is right

The full sign-class ensemble supplies a thermodynamic reference for approximate shortest-vector problems without yielding an algorithm.
Primitive-direction visibility occurs exactly at the critical inverse temperature gamma to the minus 2 rather than at c equals 1.
The Poisson point process limit for the edge window mass holds uniformly in the scaling parameter a.
Thermal concentration of primitive mass is quenched, meaning it holds for almost every realization of the random lattice.

Where Pith is reading between the lines

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

The same edge statistics may govern other multiplicative functionals of lattice vectors beyond squared length.
The critical value gamma to the minus 2 could be tested by replacing the fixed approximation factor with a slowly growing one.
The Poisson-Dirichlet limit suggests that typical approximate shortest vectors become exchangeable under the Gibbs measure for c greater than 1.

Load-bearing premise

The random lattice is treated as quenched geometric disorder realized by the Haar-random unimodular lattice in the high-dimensional limit.

What would settle it

Direct Monte-Carlo sampling of the ranked Gibbs weights on a sequence of random lattices with growing dimension n, checking whether they converge in distribution to Poisson-Dirichlet(1/c) for c greater than 1.

read the original abstract

We study Gibbs measures on high--dimensional Haar--random unimodular lattices, where the energy of a lattice vector is its squared Euclidean norm. The random lattice is viewed as quenched geometric disorder, and $c>0$ denotes the scaled inverse temperature. We first analyze the edge window of vectors whose length is within the factor $e^{a/n}$ of the shortest length, with fixed $a$ as $n\to\infty$. For the full sign--class Gibbs ensemble, we prove a Poisson point process limit theorem for the Gibbs mass of this window. The mass vanishes in probability for $0<c\le1$, while for $c>1$ it has a nontrivial Poisson limit, and the ranked Gibbs weights converge to the Poisson--Dirichlet distribution with parameter $1/c$. We then pass to a primitive--direction Gibbs ensemble and consider a fixed approximation factor $\gamma>1$. For this modified ensemble, we prove a weighted moment formula and a quenched thermal concentration result in the high--temperature range $0<c<1$. This yields the primitive fixed--factor visibility curve $c=\gamma^{-2}$ for approximate shortest directions. More precisely, the primitive Gibbs mass of the fixed--factor window tends to zero for $c<\gamma^{-2}$, to one for $\gamma^{-2}<c<1$, and to $1/2$ at the critical boundary $c=\gamma^{-2}$. Thus the fixed--factor theorem is a visibility statement for an idealized primitive target measure, not for the original full lattice Gibbs measure. The results provide a random--lattice thermodynamic reference model for Gibbs targets related to approximate shortest vectors, without implying an efficient algorithm for the shortest vector problem.

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.

Desk Editor's Note private letter to a colleague

The paper proves new Poisson point process and Poisson-Dirichlet limits plus a primitive visibility curve for Gibbs ensembles on Haar-random lattices, but the results stay inside a narrow high-dimensional reference model with no algorithmic payoff.

read the letter

The core contribution here is a set of limit theorems for Gibbs measures on high-dimensional Haar-unimodular random lattices, with energy given by squared Euclidean norm. For the full sign-class ensemble they establish Poisson point process convergence of the Gibbs mass in the edge window scaled by e^{a/n}: the mass goes to zero in probability for c ≤ 1 and has a nontrivial Poisson limit for c > 1, with the ranked weights converging to Poisson-Dirichlet(1/c). For the primitive-direction ensemble they add a weighted moment formula and quenched concentration that produces the fixed-factor visibility threshold at c = γ^{-2}, where the primitive mass jumps from near zero to near one (with value 1/2 exactly at the boundary). They are explicit that this visibility statement applies only to the idealized primitive measure, not the original full Gibbs measure.

The work is careful about its modeling choices and does not claim algorithmic consequences for the shortest-vector problem, which keeps the claims proportionate. The use of standard point-process machinery on the quenched random lattice appears internally consistent from the statements given.

The main limitation is scope: everything is proved in the n → ∞ limit under a specific scaling and for these two ensembles only. The full derivations are not reproduced here, so any gaps in the technical steps remain unchecked, but the abstract framing shows no obvious circularity or hidden fitting. The results supply a clean thermodynamic reference model rather than a broad advance.

This is for readers already working in random-lattice probability or point processes on disordered media. A specialist in that subfield would find the explicit limits and the visibility curve useful. It is worth sending to peer review because the claims are stated precisely enough to be checked and the modeling is laid out up front.

Referee Report

0 major / 1 minor

Summary. The paper analyzes Gibbs measures on high-dimensional Haar-random unimodular lattices with energy given by squared Euclidean norm, treating the lattice as quenched disorder. For the full sign-class ensemble it proves a Poisson point process limit for the Gibbs mass of the edge window (vectors within factor e^{a/n} of shortest length), showing the mass vanishes in probability for 0 < c ≤ 1, admits a nontrivial Poisson limit for c > 1, and that the ranked weights converge to the Poisson-Dirichlet distribution with parameter 1/c. For a modified primitive-direction ensemble it establishes a weighted moment formula and quenched thermal concentration in the high-temperature regime 0 < c < 1, yielding a visibility threshold at c = γ^{-2} for fixed approximation factor γ > 1; the primitive Gibbs mass tends to 0 below the threshold, to 1 above it (but still <1), and to 1/2 at criticality. The results are presented as a thermodynamic reference model only, with explicit caveats that they apply to idealized ensembles and do not address algorithmic questions for the shortest-vector problem.

Significance. If the stated limit theorems hold, the work supplies a clean, parameter-free reference model for the thermal statistics of approximate shortest vectors on random lattices. The explicit Poisson-Dirichlet and concentration statements, together with the distinction between full and primitive ensembles, give a precise picture of how inverse temperature c controls mass concentration near the shortest-length shell. This is of potential interest to random geometry and lattice-based cryptography as a theoretical benchmark, while the paper's careful qualification that the fixed-factor result concerns only the primitive measure prevents over-interpretation.

minor comments (1)

[Abstract] The abstract packs two distinct ensembles and three separate limit statements into a single paragraph; a modest reorganization (e.g., one sentence per ensemble) would improve readability without altering content.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for their positive summary, significance assessment, and recommendation to accept the manuscript. We are pleased that the distinction between the full sign-class ensemble and the primitive-direction ensemble, along with the explicit caveats regarding applicability to idealized ensembles only, was noted.

Circularity Check

0 steps flagged

No significant circularity detected

full rationale

The paper states explicit limit theorems (Poisson point process convergence of Gibbs mass, Poisson-Dirichlet convergence of ranked weights, and quenched thermal concentration) inside a precisely defined model of Haar-random unimodular lattices in high dimension n→∞ with energy equal to squared Euclidean norm and edge window scaled by e^{a/n}. Modeling choices are declared as setup assumptions upfront, results are qualified as applying only to the idealized primitive measure, and no reduction of any claimed prediction or theorem to a fitted parameter, self-definition, or load-bearing self-citation is present. The derivation chain is self-contained against the stated random-lattice and point-process framework.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

Relies on standard probabilistic assumptions for random lattices and Gibbs measures; no free parameters fitted to data, no invented entities, and axioms are background mathematical facts.

axioms (2)

standard math Existence and uniqueness of Haar measure on the space of unimodular lattices
Invoked to define the random lattice model in the setup paragraph.
standard math Convergence of point processes and Poisson-Dirichlet limits under the stated scaling
Background result used for the edge statistics theorem.

pith-pipeline@v0.9.1-grok · 5844 in / 1506 out tokens · 41863 ms · 2026-07-02T00:57:46.626025+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

17 extracted references · 15 canonical work pages

[1]

Derrida, Random–energy model: Limit of a family of disordered models, Physical Review Letters 45 (1980), no

B. Derrida, Random–energy model: Limit of a family of disordered models, Physical Review Letters 45 (1980), no. 2, 79–82.https://doi.org/10.1103/ PhysRevLett.45.79

1980
[2]

Derrida, Random–energy model: An exactly solvable model of disordered systems, Physical Review B 24 (1981), no

B. Derrida, Random–energy model: An exactly solvable model of disordered systems, Physical Review B 24 (1981), no. 5, 2613–2626.https://doi.org/ 10.1103/PhysRevB.24.2613

work page doi:10.1103/physrevb.24.2613 1981
[3]

S¨ odergren, On the Poisson distribution of lengths of lattice vectors in a random lattice, Mathematische Zeitschrift 269 (2011), no

A. S¨ odergren, On the Poisson distribution of lengths of lattice vectors in a random lattice, Mathematische Zeitschrift 269 (2011), no. 3–4, 945–954. https://doi.org/10.1007/s00209-010-0772-8

work page doi:10.1007/s00209-010-0772-8 2011
[4]

S¨ odergren, On the value distribution and moments of the Epstein zeta function to the right of the critical strip, Journal of Number Theory 131 (2011), no

A. S¨ odergren, On the value distribution and moments of the Epstein zeta function to the right of the critical strip, Journal of Number Theory 131 (2011), no. 7, 1176–1208.https://doi.org/10.1016/j.jnt.2010.12.003

work page doi:10.1016/j.jnt.2010.12.003 2011
[5]

S¨ odergren, On the value distribution of the Epstein zeta function in the critical strip, Duke Mathematical Journal 162 (2013), no

A. S¨ odergren, On the value distribution of the Epstein zeta function in the critical strip, Duke Mathematical Journal 162 (2013), no. 1, 1–48.https: //doi.org/10.1215/00127094-1903389 26

work page doi:10.1215/00127094-1903389 2013
[6]

Ajtai, Generating hard instances of lattice problems, extended abstract, in Proceedings of the 28th ACM Symposium on Theory of Computing, ACM, 1996, pp

M. Ajtai, Generating hard instances of lattice problems, extended abstract, in Proceedings of the 28th ACM Symposium on Theory of Computing, ACM, 1996, pp. 99–108.https://doi.org/10.1145/237814.237838

work page doi:10.1145/237814.237838 1996
[7]

Micciancio and S

D. Micciancio and S. Goldwasser,Complexity of Lattice Problems: A Cryptographic Perspective, Springer, 2002.https://doi.org/10.1007/ 978-1-4615-0897-7

2002
[8]

Regev, On lattices, learning with errors, random linear codes, and cryp- tography, Journal of the ACM 56 (2009), no

O. Regev, On lattices, learning with errors, random linear codes, and cryp- tography, Journal of the ACM 56 (2009), no. 6, Article 34, 40 pp.https: //doi.org/10.1145/1568318.1568324

work page doi:10.1145/1568318.1568324 2009
[9]

Micciancio and O

D. Micciancio and O. Regev, Worst–case to average–case reductions based on Gaussian measures, SIAM Journal on Computing 37 (2007), no. 1, 267–302. https://doi.org/10.1137/S0097539705447360

work page doi:10.1137/s0097539705447360 2007
[10]

C. L. Siegel, A mean value theorem in geometry of numbers, Annals of Math- ematics 46 (1945), no. 2, 340–347.https://doi.org/10.2307/1969027

work page doi:10.2307/1969027 1945
[11]

J. F. C. Kingman, Random discrete distributions, Journal of the Royal Sta- tistical Society. Series B (Methodological) 37 (1975), no. 1, 1–15.https: //doi.org/10.1111/j.2517-6161.1975.tb01024.x

work page doi:10.1111/j.2517-6161.1975.tb01024.x 1975
[12]

Pitman and M

J. Pitman and M. Yor, The two–parameter Poisson–Dirichlet distribution derived from a stable subordinator, Annals of Probability 25 (1997), no. 2, 855–900.https://doi.org/10.1214/aop/1024404422

work page doi:10.1214/aop/1024404422 1997
[13]

A. M. Macbeath and C. A. Rogers, A modified form of Siegel’s mean–value theorem, Proceedings of the Cambridge Philosophical Society 51 (1955), no. 4, 565–576.https://doi.org/10.1017/S0305004100030656

work page doi:10.1017/s0305004100030656 1955
[14]

A. M. Macbeath and C. A. Rogers, A modified form of Siegel’s mean–value theorem. II, Proceedings of the Cambridge Philosophical Society 54 (1958), no. 3, 322–326.https://doi.org/10.1017/S030500410003351X

work page doi:10.1017/s030500410003351x 1958
[15]

L. K. Grover, A fast quantum mechanical algorithm for database search, in Proceedings of the 28th ACM Symposium on Theory of Computing, ACM, 1996, pp. 212–219.https://doi.org/10.1145/237814.237866

work page doi:10.1145/237814.237866 1996
[16]

Brassard, P

G. Brassard, P. Høyer, M. Mosca, and A. Tapp, Quantum amplitude amplifi- cation and estimation, inQuantum Computation and Information, Contem- porary Mathematics 305, American Mathematical Society, 2002, pp. 53–74. https://doi.org/10.1090/conm/305/05215

work page doi:10.1090/conm/305/05215 2002
[17]

Temme, T

K. Temme, T. J. Osborne, K. G. H. Vollbrecht, D. Poulin, and F. Verstraete, Quantum Metropolis sampling, Nature 471 (2011), no. 7336, 87–90.https: //doi.org/10.1038/nature09770 27

work page doi:10.1038/nature09770 2011

[1] [1]

Derrida, Random–energy model: Limit of a family of disordered models, Physical Review Letters 45 (1980), no

B. Derrida, Random–energy model: Limit of a family of disordered models, Physical Review Letters 45 (1980), no. 2, 79–82.https://doi.org/10.1103/ PhysRevLett.45.79

1980

[2] [2]

Derrida, Random–energy model: An exactly solvable model of disordered systems, Physical Review B 24 (1981), no

B. Derrida, Random–energy model: An exactly solvable model of disordered systems, Physical Review B 24 (1981), no. 5, 2613–2626.https://doi.org/ 10.1103/PhysRevB.24.2613

work page doi:10.1103/physrevb.24.2613 1981

[3] [3]

S¨ odergren, On the Poisson distribution of lengths of lattice vectors in a random lattice, Mathematische Zeitschrift 269 (2011), no

A. S¨ odergren, On the Poisson distribution of lengths of lattice vectors in a random lattice, Mathematische Zeitschrift 269 (2011), no. 3–4, 945–954. https://doi.org/10.1007/s00209-010-0772-8

work page doi:10.1007/s00209-010-0772-8 2011

[4] [4]

S¨ odergren, On the value distribution and moments of the Epstein zeta function to the right of the critical strip, Journal of Number Theory 131 (2011), no

A. S¨ odergren, On the value distribution and moments of the Epstein zeta function to the right of the critical strip, Journal of Number Theory 131 (2011), no. 7, 1176–1208.https://doi.org/10.1016/j.jnt.2010.12.003

work page doi:10.1016/j.jnt.2010.12.003 2011

[5] [5]

S¨ odergren, On the value distribution of the Epstein zeta function in the critical strip, Duke Mathematical Journal 162 (2013), no

A. S¨ odergren, On the value distribution of the Epstein zeta function in the critical strip, Duke Mathematical Journal 162 (2013), no. 1, 1–48.https: //doi.org/10.1215/00127094-1903389 26

work page doi:10.1215/00127094-1903389 2013

[6] [6]

Ajtai, Generating hard instances of lattice problems, extended abstract, in Proceedings of the 28th ACM Symposium on Theory of Computing, ACM, 1996, pp

M. Ajtai, Generating hard instances of lattice problems, extended abstract, in Proceedings of the 28th ACM Symposium on Theory of Computing, ACM, 1996, pp. 99–108.https://doi.org/10.1145/237814.237838

work page doi:10.1145/237814.237838 1996

[7] [7]

Micciancio and S

D. Micciancio and S. Goldwasser,Complexity of Lattice Problems: A Cryptographic Perspective, Springer, 2002.https://doi.org/10.1007/ 978-1-4615-0897-7

2002

[8] [8]

Regev, On lattices, learning with errors, random linear codes, and cryp- tography, Journal of the ACM 56 (2009), no

O. Regev, On lattices, learning with errors, random linear codes, and cryp- tography, Journal of the ACM 56 (2009), no. 6, Article 34, 40 pp.https: //doi.org/10.1145/1568318.1568324

work page doi:10.1145/1568318.1568324 2009

[9] [9]

Micciancio and O

D. Micciancio and O. Regev, Worst–case to average–case reductions based on Gaussian measures, SIAM Journal on Computing 37 (2007), no. 1, 267–302. https://doi.org/10.1137/S0097539705447360

work page doi:10.1137/s0097539705447360 2007

[10] [10]

C. L. Siegel, A mean value theorem in geometry of numbers, Annals of Math- ematics 46 (1945), no. 2, 340–347.https://doi.org/10.2307/1969027

work page doi:10.2307/1969027 1945

[11] [11]

J. F. C. Kingman, Random discrete distributions, Journal of the Royal Sta- tistical Society. Series B (Methodological) 37 (1975), no. 1, 1–15.https: //doi.org/10.1111/j.2517-6161.1975.tb01024.x

work page doi:10.1111/j.2517-6161.1975.tb01024.x 1975

[12] [12]

Pitman and M

J. Pitman and M. Yor, The two–parameter Poisson–Dirichlet distribution derived from a stable subordinator, Annals of Probability 25 (1997), no. 2, 855–900.https://doi.org/10.1214/aop/1024404422

work page doi:10.1214/aop/1024404422 1997

[13] [13]

A. M. Macbeath and C. A. Rogers, A modified form of Siegel’s mean–value theorem, Proceedings of the Cambridge Philosophical Society 51 (1955), no. 4, 565–576.https://doi.org/10.1017/S0305004100030656

work page doi:10.1017/s0305004100030656 1955

[14] [14]

A. M. Macbeath and C. A. Rogers, A modified form of Siegel’s mean–value theorem. II, Proceedings of the Cambridge Philosophical Society 54 (1958), no. 3, 322–326.https://doi.org/10.1017/S030500410003351X

work page doi:10.1017/s030500410003351x 1958

[15] [15]

L. K. Grover, A fast quantum mechanical algorithm for database search, in Proceedings of the 28th ACM Symposium on Theory of Computing, ACM, 1996, pp. 212–219.https://doi.org/10.1145/237814.237866

work page doi:10.1145/237814.237866 1996

[16] [16]

Brassard, P

G. Brassard, P. Høyer, M. Mosca, and A. Tapp, Quantum amplitude amplifi- cation and estimation, inQuantum Computation and Information, Contem- porary Mathematics 305, American Mathematical Society, 2002, pp. 53–74. https://doi.org/10.1090/conm/305/05215

work page doi:10.1090/conm/305/05215 2002

[17] [17]

Temme, T

K. Temme, T. J. Osborne, K. G. H. Vollbrecht, D. Poulin, and F. Verstraete, Quantum Metropolis sampling, Nature 471 (2011), no. 7336, 87–90.https: //doi.org/10.1038/nature09770 27

work page doi:10.1038/nature09770 2011