WorksMathematics
Large deviations and the asymptotic geometry of Lorentz balls
Abstract
We study the uniform distribution on finite-dimensional Lorentz balls in the range . We prove a Sanov-type large deviation principle for the empirical measures of the absolute coordinates of a random vector uniformly distributed on . The rate function is given by an entropy deficit under a weighted quantile constraint. We show that the entropy maximizer is unique and characterize its distribution function as the unique increasing solution of a nonlinear boundary-value problem. The limiting distribution has compact support, for which we obtain an explicit bound. These results establish the conjectured empirical limit of Kabluchko, Prochno and Sonnleitner [Journal of Functional Analysis, 288(1):110682, 2025] in the range . As consequences, we determine the exponential volume asymptotics of Lorentz balls and obtain asymptotic independence of any fixed number of coordinates. We also prove a law of large numbers for Lorentz norms with finite parameters and use it, together with the volume asymptotics, to obtain a Schechtman–Schmuckenschläger type theorem for intersections of volume-normalized Lorentz balls. We further derive a variational formula for the logarithmic intersection asymptotics. The proof combines an exponential change of measure with quantile interpolation and a variational argument that does not assume regularity of the maximizing quantile in advance.
Introduction and main results
The probabilistic study of high-dimensional convex bodies is a central part of asymptotic geometric analysis. Endowing a convex body with normalized Lebesgue measure connects its geometric properties to the distribution of a random vector and makes probabilistic limit theorems available for their study. The unit balls of the classical sequence spaces provide a particularly instructive setting. A fundamental tool is the probabilistic representation obtained by Schechtman and Zinn [21] and independently by Rachev and Rüschendorf [19]: a uniformly distributed point can be represented using independent -generalized Gaussian random variables, normalization by their -norm, and an independent radial factor. This representation underlies the analysis of coordinate distributions, norms, and asymptotic intersection volumes for these balls.
An important impetus for the systematic study of large deviations in this setting came from Gantert, Kim and Ramanan [8], who established annealed and quenched large deviation principles for one-dimensional projections of uniformly distributed points in -balls. The dependence of the speeds and rate functions on reveals geometric information beyond the Gaussian fluctuation regime. Alonso-Gutiérrez, Prochno and Thäle [1] subsequently studied the Euclidean norms of projections onto random subspaces whose dimensions may grow with the ambient dimension. Kabluchko, Prochno and Thäle [13] obtained central limit theorems and large deviation principles for -norms of random points in -balls, and later complemented these results by moderate deviation principles [16], describing the intermediate scale between central limit and large deviation behaviour.
A closely related development concerns the empirical distribution of coordinates. Kim and Ramanan [17] proved Sanov-type large deviation principles for empirical measures under the cone and surface probability measures on high-dimensional -spheres and used them to derive conditional limit theorems. In the matrix setting, Kabluchko, Prochno and Thäle [15] established corresponding large deviation principles for empirical eigenvalue measures of self-adjoint matrix balls and singular-value measures of Schatten balls. Such results identify limiting empirical distributions through variational problems for the rate function. For an account of these developments and their connections to geometric functional analysis, we refer to the survey [18].
The extension of this programme to Orlicz balls brought maximum entropy methods and Gibbs measures to the fore. Kabluchko and Prochno [11] related the uniform distribution on an Orlicz ball to Gibbs distributions with potential given by the defining Orlicz function. This provided an alternative to the explicit probabilistic representation for -balls and led to asymptotic volume formulas (see also [2]) and results on intersection volumes. Building on this approach, Frühwirth and Prochno [6] established Sanov-type large deviation principles for the empirical measures of the coordinates, together with conditional limit theorems. These results extend the connection between empirical distributions, entropy maximization, and asymptotic geometry beyond the classical power constraints.
Lorentz balls form another natural extension of the classical -balls. For , their defining constraint depends on the order of the absolute coordinates and therefore cannot be written as a sum of one-coordinate potentials. To make this precise, let denote the decreasing rearrangement of . For , the finite-dimensional Lorentz space is equipped with the norm
and we denote its unit ball by
When , this is the usual -ball. The dependence on the decreasing rearrangement is the main additional obstacle when .
Kabluchko, Prochno and Sonnleitner [12] developed a probabilistic approach to Lorentz balls in the case . They obtained an explicit probabilistic representation of the uniform distribution, identified the limiting empirical law, and derived coordinate and volumetric limit theorems, including a Schechtman–Schmuckenschläger type result for intersections with classical -balls. For general , they formulated a conjecture describing the limiting distribution through a nonlinear differential equation with boundary conditions. Their heuristic derivation uses a weighted quantile constraint and a maximum entropy principle.
In this paper, we establish this conjectured description in the range . We first prove a large deviation principle for the empirical measures on the positive orthant. We then show that the associated entropy maximization problem has a unique solution and derive the differential equation together with its boundary conditions. A converse argument proves uniqueness of the admissible solution of the boundary-value problem. In particular, the differential equation characterizes the limiting law. The same analysis determines the exponential volume asymptotics and yields a Poincaré–Maxwell–Borel principle for Lorentz balls. We also prove a law of large numbers for Lorentz norms with parameters . Together with the volume asymptotics, this yields a Schechtman–Schmuckenschläger type theorem for intersections of two volume-normalized Lorentz balls. We further determine the exponential decay rate of the intersection volumes below the threshold.
Main results
Throughout the paper, unless stated otherwise, we fix and use the normalization
Let , let be uniformly distributed on , and put
We equip with the topology of weak convergence. For , its increasing quantile function is
For , define
Every measure in has th moment at most one. We write for Lebesgue measure on and for its restriction to a Borel set . For a measure and a measurable map defined on its underlying space, we write for the pushforward of under , defined by .
for every measurable set in the target space. For a probability measure with finite first moment, we write
and put . This entropy is well defined in . Let us define
Theorem A (Large deviations and volume asymptotics). Let . Then , and the empirical measures satisfy a large deviation principle on with speed and good rate function
Moreover,
Thus the exponential volume asymptotics are determined by a one-dimensional variational problem. Our second theorem identifies its maximizer and the corresponding empirical limit.
Theorem B (The limiting distribution). Let .
(i) There is a unique measure with , and . Its quantile extends to , with , and satisfies
(ii) The support of is , where
Its distribution function belongs to and is the unique increasing solution, with positive derivative in the interior, of
Here uniqueness is among pairs with , and , on ; the solution necessarily has . Derivatives at finite endpoints are understood one-sidedly.
(iii) The empirical measures converge to in probability with respect to the topology of weak convergence. If is uniformly distributed on the full scaled ball , then
The measure has the continuous density

Figure 1. Numerical limiting densities of the coordinate distributions on high-dimensional Lorentz balls. The left panel fixes and varies ; the right panel fixes and varies . Each density is symmetric, has total mass one, and vanishes outside a compact interval.
Theorem B proves the empirical-limit conjecture in [12] for . The cases and are discussed in that paper and correspond, respectively, to its explicit Lorentz limit and to the classical -Gaussian law. The compact support in (7) is a feature of the strict inequality . Our assertion of uniqueness in (8) includes the terminal derivative condition; the differential equation with alone does not determine a unique solution.
The permutation symmetry of turns convergence of the empirical measures into convergence of any fixed number of coordinates.
Corollary 1.1 (A Poincaré–Maxwell–Borel principle). Let be fixed and let be uniformly distributed on . Then
We also obtain an intersection-volume result in the sense of Schechtman and Schmuckenschläger [20]. It complements [7, 9, 10, 11, 14], extends the intersection result of [12] to and allows a second Lorentz ball in place of a classical -ball. For , write
If , let denote the entropy maximum for the pair , so that . Theorem A, the volume asymptotics in [12], and the classical -ball volume formula give, respectively,
Corollary 1.2 (Intersection volumes). Let and , and let be the quantile from Theorem B. Define
Then, for every with ,
For every fixed , there exists such that
Moreover, . If , then if and only if . In particular, the intersection volume at tends to zero for two distinct parameter pairs in the strict parameter range.
For , the threshold can equivalently be written as . For a classical -ball, corresponding to , it is
The variational characterization of is given in (55), and the exponential asymptotics are proved in Proposition 6.4.
Outline of the proofs
The first step is to express the Lorentz constraint as an integral against the empirical quantile. The resulting weight depends on , so the admissible sets vary with the dimension. We control these sets inside a common weakly compact moment ball and identify all their possible weak limits. An exponential change of measure then relates restricted volumes to the empirical measures of independent exponential random variables. Sanov’s theorem gives the upper bound after truncation of the first-moment tilt. For the lower bound, we use boundedly supported comparison measures and approximate arbitrary finite-entropy admissible measures by truncation and dilation.
The variational analysis is carried out in quantile coordinates. Concavity of the entropy along quantile interpolation and strict convexity of the energy give uniqueness. Before taking variations, we remove any initial offset and singular part of the maximizing quantile. Multiplicative perturbations of its derivative then yield (6) with an explicit integrable majorant. The integral equation supplies the required regularity, the boundary conditions, and the compact-support estimate. Finally, an entropy comparison shows that every admissible solution of the boundary-value problem is the same maximizer.
For the intersection results, convexity yields uniform exponential tails for the one-coordinate marginals. This strengthens empirical convergence to Wasserstein convergence and gives a law of large numbers for every Lorentz norm with finite parameters. Combining this with the volume asymptotics yields the threshold. A second quantile constraint in the restricted-volume argument gives the logarithmic intersection asymptotics.
Section 2 contains the quantile representation and the compactness and entropy estimates. Section 3 proves Theorem A. Section 4 establishes uniqueness of the entropy maximizer and the empirical and coordinate limits. Section 5 completes the characterization of the limiting distribution. Section 6 proves the intersection-volume results.
Quantiles, compactness and entropy
We write for weak convergence of probability measures. The space is Polish for this topology; throughout, denotes any compatible metric. We use the facts that weak convergence implies at every continuity point of , hence for almost every , and that
In particular, integration against can be written as integration of a function of on . We refer to [3 Appendix D] for the weak-topology facts used below.
The discrete constraint
For , we define
We recall the finite-dimensional quantile identity from [12] and include its short proof to fix the mesh convention.
Lemma 2.1. For , let . Then
Moreover, for any ,
Proof. On , the increasing quantile of equals . This remains true when coordinates coincide. Integrating over these intervals and replacing by gives (2.3). Since , is increasing and integrable, and takes its value at the left endpoint of each interval. This proves (2.4).
A common compact set
Let
The inequality implies that all the sets and are contained in .
Lemma 2.2. The set is weakly compact and is continuous on . The functionals and are weakly lower semicontinuous. If , then
Consequently, is compact and, for every ,
Proof. The th-moment functional is lower semicontinuous, as it is the supremum of the continuous functionals . Thus is closed. The bound holds uniformly on , so Prohorov’s theorem gives compactness. Furthermore,
Hence the bounded continuous functionals converge uniformly to on . If , quantile convergence and (20) allow us to apply Fatou’s lemma to , giving (21). The same argument with a fixed weight or proves lower semicontinuity. In particular, all these functionals are Borel measurable and is a closed subset of and hence compact.
If (22) were false, there would be and with and . A subsequence converges to some . The proof of (21), applied along this subsequence, gives . Hence , contradicting .
Entropy
Let be the distribution. We use the relative entropy
It is nonnegative and weakly lower semicontinuous; see [3 Section 6.2].
For every probability measure with finite first moment, the standard entropy identity gives
with both sides understood in the extended sense. Since is continuous on by Lemma 2.2, the weak lower semicontinuity of relative entropy implies that is upper semicontinuous on . Moreover, for . On the other hand, the uniform distribution on , with , belongs to and has entropy . Consequently, .
Restricted volumes and large deviations
For a Borel set , define
This is well defined because is weakly continuous. Indeed, is continuous for every .
Proposition 3.1 (Restricted-volume bounds). For every closed and every open
we use the conventions .
Proof. The upper bound. Let be independent with law and put . Since on the positive orthant, Lemma 2.1 gives the exact identity
By Sanov’s theorem [3 Theorem 6.2.10], the sequence satisfies an LDP for weak convergence, with speed and good rate function . Fix and set . This set is compact. By (2.6) and (2.7), for all sufficiently large ,
The restricted Laplace upper bound for a bounded continuous functional on a closed set [4 Appendix B.2, pp. 321–322] therefore yields
After letting , we consider the compact sets , which decrease to . If this intersection is nonempty, upper semicontinuity of and compactness imply
For completeness, choose a maximizer in each and pass to a convergent subsequence in ; its limit lies in , and upper semicontinuity gives the nontrivial inequality. This argument also applies when the supremum is . If is empty, compactness implies that is empty for some . This proves (3.1).
The lower bound for bounded support and strict feasibility. Let have support in , finite entropy, and . For independent random variables with law , their empirical measures converge weakly to almost surely. This follows, for example, by applying the strong law to the indicators of intervals with rational endpoints and using monotonicity of distribution functions. Quantile convergence and the bound
give almost surely by dominated convergence. Moreover, finite entropy implies , so
Fix . Let consist of the vectors for which for all and
Then , and
Taking logarithms and then letting shows that
Approximation of an arbitrary admissible measure. Let have finite entropy. For , let be its conditional law on , with density . Then , and
Consequently, and . Also,
Here , and by finite entropy.
For , write for the pushforward of by . Directly from the density and quantile formulas,
For large and then sufficiently close to one, the measure belongs to , has bounded support and energy strictly below one, and has entropy arbitrarily close to . Apply (3.4) and take the supremum over the finite-entropy measures in . If there are no such measures, the lower bound is vacuous. This proves (3.2).
Proof of Theorem A. Apply Proposition 3.1 to the whole space to obtain
Since is uniform on ,
Subtracting (3.7) from the restricted-volume bounds gives the LDP upper bound for closed sets and lower bound for open sets, with rate function (1.4). The rate function is nonnegative. For every finite ,
is a closed subset of the compact set , by upper semicontinuity of . Thus is a good rate function.
Finally, orthant symmetry and (1.1) give
Taking th roots and using (3.7) proves (1.5).
The entropy maximizer and empirical limits
Existence of a maximizer follows from compactness of and upper semicontinuity of the entropy. To prove uniqueness, we interpolate between quantiles. This requires an entropy formula that remains valid when the quantile has a singular part.
Entropy in quantile coordinates
Lemma 4.1 (Quantile entropy formula). Let be nondecreasing and let . Then
If these conditions hold and has finite first moment, then
where both sides are allowed to take the value . In particular, if , then .
Proof. Suppose , and let be its distribution function. Since has no atoms, is strictly increasing: an interval on which is constant would create an atom. The pushforward identity therefore gives for . The equality holds outside a Lebesgue-null set. Its preimage under has measure zero because . Since is differentiable with finite derivative almost everywhere, the chain rule gives
Here is positive and finite almost everywhere: the sets where or have -measure zero after choosing a finite almost everywhere version of the density. This proves almost everywhere.
Conversely, assume almost everywhere. Let be a Lebesgue-null Borel set, and let be the set of points in at which exists and lies in . Then is null. The restriction is injective and differentiable relative to . The one-dimensional area formula on a measurable domain [5 Theorem 263D(iv)] gives
Thus is null and , proving absolute continuity of . This argument does not assume absolute continuity of .
Under these equivalent conditions, (33) holds. Hence almost everywhere. Apply the pushforward identity to the positive and negative parts separately. If has finite first moment, (2.8) ensures that the entropy is well defined, and the resulting equality is (4.1).
Proposition 4.2 (Existence and uniqueness). There is a unique entropy maximizer in , and .
Proof. Lemma 2.2, upper semicontinuity of , and give existence and finite maximizing entropy. Every maximizer has energy one: otherwise, a dilation by some would still belong to and would increase the entropy by , by (3.6).
Suppose that are two maximizers, with quantiles . For , define
The function is nondecreasing and left-continuous, so it is the quantile of . Since and differ on a set of positive measure, strict convexity of and positivity of give
By Lemma 4.1, almost everywhere, and hence almost everywhere. Since , it has finite first moment. Thus Lemma 4.1 and concavity of the logarithm yield
In view of (4.3), dilating slightly gives an admissible measure with entropy greater than , a contradiction.
Empirical measures and fixed coordinates
Proof of Theorem B(iii). Fix and put . By compactness, upper semicontinuity, and uniqueness of the maximizer,
with the same conclusion if the intersection is empty. Theorem A gives
This proves in probability.
To pass to the full scaled ball, use its invariance under coordinatewise sign changes. We may represent
where the are independent symmetric signs, independent of . Orthant boundaries have Lebesgue measure zero, so this representation is exact in law. For , put for . Conditional on , the centered sign fluctuation
has mean zero and variance at most . Hence in probability, whereas
Finite intersections of neighborhoods specified by bounded continuous test functions form a neighborhood basis for weak convergence. Thus convergence of these integrals in probability gives (1.9). The density formula (1.10) follows from the regularity established in Section 5.
Proof of Corollary 1.1. Let and fix . Theorem B and boundedness give
For , expanding the product gives
Write . By exchangeability, every term with distinct indices has expectation . The total contribution of repeated indices therefore gives the estimate
The one-coordinate marginals converge to and are tight, hence the -coordinate laws are tight as well. Every subsequential limit has the displayed product integrals of , which determine this product measure. This proves the claim.
The variational equation and compact support
We now identify the entropy maximizer. The first step justifies working with the derivative of its quantile.
Lemma 5.1. Let . Then almost everywhere on , and for every ,
In particular, after setting , the function is absolutely continuous on for every .
Proof. By Lemma 4.1, almost everywhere. Monotonicity gives
Consequently, is finite, satisfies , and has derivative almost everywhere. Since is continuous and strictly increasing, it is the quantile function of its induced law . The inequality implies , and the quantile entropy formula gives
Uniqueness of the maximizer gives . Since and are the corresponding quantiles, they agree on . This removes any initial offset, jumps, and singular continuous part of .
Admissible variations
For a quantile , write and for the energy and entropy of its induced law. If and , scaling gives
Since for , the functional
therefore satisfies
This observation replaces the constrained problem by an unconstrained one without introducing an unknown Lagrange multiplier.
Proposition 5.2. The maximizing quantile belongs to and satisfies (1.6). If , then satisfies
Proof. Write and use Lemma 5.1. Let be bounded and measurable. For sufficiently close to zero, define
Put and assume . Then almost everywhere and
Thus induces a measure with finite positive energy, hence with finite first moment. Moreover,
so it has finite entropy by Lemma 4.1. Both positive and negative small are admissible in (5.3).
Set , so and . Differentiation under the integral sign is justified by the bounds
The second majorant is integrable because . Since has a maximum at zero, its derivative there vanishes:
The double integral is absolutely integrable, since
Fubini’s theorem turns (5.5) into
Because is arbitrary,
We now show that and that (1.6) holds for every . Hölder’s inequality gives
Hence is continuous on . Since for , we have for , whereas . By (5.1) and (5.6),
This proves and (1.6). On ,
The inverse function theorem now yields , and therefore
This proves (5.4).
The support and endpoint conditions
Proposition 5.3. The support of is , with satisfying (7). Its distribution function satisfies
Moreover, , and its density is strictly decreasing on .
Proof. By monotonicity of , for ,
Since , it follows that
Integrating from zero and using gives
Thus and (7) holds.
The continuous strictly increasing quantile maps onto . Its induced law therefore has support , and the inverse extends continuously with values zero and one at the endpoints. Since in the interior and is continuous on ,
The one-sided endpoint derivatives have these values by the mean value theorem. This proves the stated regularity. Finally, (39) has a strictly negative right-hand side on , so is strictly decreasing there.
Uniqueness of the boundary-value solution
The preceding results give an admissible solution of (8). To obtain a characterization, we must also show that every such solution yields the entropy maximizer. The following argument includes competitors whose quantiles have a singular part.
Proposition 5.4 (Converse characterization). Let satisfy the regularity, positivity and boundary assumptions in Theorem B(ii), as well as (8). Then and is the distribution function of .
Proof. Let on and set . The differential equation gives
Integrating from to and letting , the terminal condition on implies
No finite-energy assumption is needed for this identity: the improper integral equals the finite quantity . We have . The same monotonicity estimate as in Proposition 5.3, integrated first from to and then with , yields (5.8). Consequently,
In particular, is finite and positive, and (5.9) implies with . Therefore . Tonelli’s theorem now gives the normalization
The law induced by also has finite entropy. Indeed, , is bounded on , and for ,
Thus near one, and . The quantile entropy formula from Lemma 4.1 applies.
Let be the quantile of any with finite entropy. By monotonicity,
whether or not is absolutely continuous. Using Tonelli, Hölder, and (5.10), we obtain
Finally, gives
Both logarithms are integrable, and the right-hand side is finite by the preceding estimate, so every integral in this comparison is justified. Measures of entropy do not affect maximality. Hence the law induced by is an entropy maximizer in , and Proposition 4.2 identifies it with . Its quantile determines and uniquely. □
Completion of the proof of Theorem B. Proposition 4.2 gives the unique maximizer and energy normalization. Proposition 5.2 gives (1.6) and the differential equation. Propositions 5.3 and 5.4 give the support estimate, boundary regularity, and uniqueness of the boundary-value solution. The empirical convergence was proved in § 4. The continuous density in (1.10) follows from and . □
Intersection volumes
Throughout this section, are the fixed source parameters and are the target parameters. We use the volume normalization and constants from (11)–(12). For , set
The corresponding discrete weight and energy are
As in Lemma 2.1,
Also, and pointwise on .
Moment bounds and Lorentz norms
We first strengthen the empirical convergence in Theorem B. For , we use the one-dimensional quantile representation of the -Wasserstein distance,
for probability measures with finite th moments.
Lemma 6.1. There is a constant such that
Consequently, for every ,
Proof. Permutation symmetry and the bound on give
Put and . Since is convex,
For , the Brunn–Minkowski inequality, followed by the weighted arithmetic–geometric mean inequality, therefore yields
The resulting log-concavity inequality is automatic if either value is zero. Thus is log-concave. Since and by Markov’s inequality, log-concavity gives for , with . This proves (50) for all . In particular,
Fix and let . Write and . The natural coupling and (6.6) imply
For fixed , Theorem B gives weakly in probability. On , weak convergence implies -convergence by almost everywhere quantile convergence and dominated convergence. Hence in probability. Moreover, as , since has compact support. Apply the triangle inequality, use Markov’s inequality for the first term, and then let to obtain (51). □
Proposition 6.2 (A law of large numbers for Lorentz norms). Let and let be uniformly distributed on . With as in (1.13), we have and
Proof. Define
We have and pointwise convergence on . Choose . Since , Minkowski’s and Hölder’s inequalities give, for measures with finite th moments,
The constant is independent of . Dominated convergence gives , because is bounded and is integrable. Thus Lemma 6.1 yields in probability. By (6.3), . The absolute coordinates of have the law of , proving (6.7). Positivity and finiteness of follow from for , boundedness of , and integrability of . □
The intersection threshold
Proof of Corollary 1.2. Put
Then and . If is uniformly distributed on , then is uniform on . Since the latter body has volume one,
Proposition 6.2 now proves (1.15) with the threshold (1.14). The bound follows from
if the threshold were smaller than one, choosing strictly between it and one would contradict (1.15).
Suppose now that , let be the entropy maximizer for the target pair, and write . The law of , where , has target energy one and entropy . Hence
If , equality holds, and uniqueness of the target maximizer implies . Thus its distribution function satisfies . Comparing the differential equations in Theorem B for the two parameter pairs and cancelling the positive derivative gives
on the interior of the target support. Letting first forces and then ; the same identity then gives . Conversely, if , the equality yields and . The assertion for follows. The exponential asymptotics follow from Proposition 6.4 below.
Remark 6.3. Corollary 1.2 does not specify the limit at . A nontrivial critical limit requires control of the norm fluctuations and of the finer volume asymptotics. For identical parameter pairs, the threshold equals one and
so the critical intersection volume equals one in every dimension.
Logarithmic intersection volumes
For , we define the following three quantities
The following proposition identifies the variational quantity with the exponential decay rate appearing in Corollary 1.2.
Proposition 6.4 (Logarithmic intersection volumes). For every , the number is finite and nonnegative, and
Moreover, for , whereas for .
Proof. Fix and put
Define
Orthant symmetry and the definitions of and give
We prove that
The upper bound. Since and , the quantile convergence and Fatou arguments in Lemma 2.2 show that and are weakly lower semicontinuous and that
In particular, all constraints above are Borel measurable and is compact. For , put
Then for all sufficiently large . Indeed, any sequence contradicting this inclusion has a convergent subsequence in ; the two moving lower bounds and place its limit in , a contradiction.
Let be independent with law and write for their empirical measures. The change of measure in (27) gives
Consequently, for fixed and all sufficiently large ,
Sanov’s theorem and the restricted Laplace upper bound used in the proof of Proposition 3.1 yield
Let and then . The compact sets decrease to , so upper semicontinuity of gives the upper bound in (58).
The lower bound. First let have bounded support, finite entropy, and
For independent with law , we have almost surely. Quantile convergence and dominated convergence give
Here both weights and are integrable. Thus eventually almost surely. Moreover, , and the strong law gives almost surely. The density comparison from the lower bound in Proposition 3.1 now gives, for every ,
The probability tends to one, hence .
For arbitrary with finite entropy, let whenever . As in the proof of Proposition 3.1, and . For , the boundedly supported measure satisfies
Apply the preceding lower bound, let and , and take the supremum over of finite entropy. Such measures exist: the uniform law on a sufficiently short interval belongs to and has entropy . Since , its entropy supremum is finite. This proves (6.12). Combining (3.7) and (6.11) now proves (6.10).
The set is compact, and the entropy is upper semicontinuous there. Since its supremum is finite and attained, uniqueness of the entropy maximizer on shows that if and only if . By (1.13) and (1.14), this is equivalent to . This also proves strict positivity of below the threshold.
In particular, if and , then Corollary 1.2 gives and
Acknowledgement
The authors were supported by the German Research Foundation under DFG project 516672205 Limit theorems for the volume of random projections of -balls. A considerable part of this work was carried out during a research semester of JP at Graz University of Technology. JP gratefully acknowledges the excellent working conditions and the hospitality provided during his stay. This work is intended to form part of SM’s doctoral dissertation.
Declaration of AI use
During the preparation of this work, the authors used Claude (Anthropic) and ChatGPT (OpenAI) as tools for literature search as well as technical and editorial assistance. It was also used to produce Figure 1. The mathematical framework and proof strategy were fully developed by the authors (and partly rely on [12]).
References
- [1]D. Alonso-Gutiérrez, J. Prochno, and C. Thäle. Large deviations for high-dimensional random projections of ℓ_p^n-balls. Advances in Applied Mathematics, 99:1–35, 2018.
- [2]F. Barthe and P. Wolff. Volume properties of high-dimensional Orlicz balls. In High dimensional probability IX—the ethereal volume, volume 80 of Progr. Probab., pages 75–95. Birkhäuser/Springer, Cham, 2023.arxiv.org/abs/2106.01675
- [3]A. Dembo and O. Zeitouni. Large Deviations Techniques and Applications, volume 38 of Applications of Mathematics. Springer, New York, second edition, 1998.
- [4]R. S. Ellis. Entropy, Large Deviations, and Statistical Mechanics. Classics in Mathematics. Springer, Berlin, 2006. Reprint of the 1985 edition.
- [5]D. H. Fremlin. Measure Theory. Vol. 2: Broad Foundations. Torres Fremlin, Colchester, second edition, 2010.
- [6]L. Frühwirth and J. Prochno. Sanov-type large deviations and conditional limit theorems for high-dimensional Orlicz balls. Journal of Mathematical Analysis and Applications, 536(1):128169, 2024.arxiv.org/abs/2111.04691
- [7]L. Frühwirth and J. Prochno. Sharp concentration phenomena in high-dimensional Orlicz balls. J. Funct. Anal., 290(6):Paper No. 111322, 45, 2026.arxiv.org/abs/2407.15579
- [8]N. Gantert, S. S. Kim, and K. Ramanan. Large deviations for random projections of ℓ^p balls. The Annals of Probability, 45(6B):4419–4476, 2017.arxiv.org/abs/1512.04988
- [9]M. Juhos, Z. Kabluchko, and J. Prochno. Limit theorems for mixed-norm sequence spaces with applications to volume distribution. Electron. J. Probab., 29:Paper No. 94, 44, 2024.DOI
- [10]M. Juhos and J. Prochno. Spectral flatness and the volume of intersections of p-ellipsoids. J. Complexity, 70:Paper No. 101617, 21, 2022.DOI
- [11]Z. Kabluchko and J. Prochno. The maximum entropy principle and volumetric properties of Orlicz balls. Journal of Mathematical Analysis and Applications, 495(1):124687, 2021.arxiv.org/abs/2007.05247
- [12]Z. Kabluchko, J. Prochno, and M. Sonnleitner. A probabilistic approach to Lorentz balls ℓ^n_{q,1}. Journal of Functional Analysis, 288(1):110682, 2025.
- [13]Z. Kabluchko, J. Prochno, and C. Thäle. High-dimensional limit theorems for random vectors in ℓ_p^n-balls. Communications in Contemporary Mathematics, 21(1):1750092, 2019.
- [14]Z. Kabluchko, J. Prochno, and C. Thäle. Intersection of unit balls in classical matrix ensembles. Israel J. Math., 239(1):129–172, 2020.arxiv.org/abs/1804.03466
- [15]Z. Kabluchko, J. Prochno, and C. Thäle. Sanov-type large deviations in Schatten classes. Annales de l’Institut Henri Poincaré, Probabilités et Statistiques, 56(2):928–953, 2020.arxiv.org/abs/1808.04862
- [16]Z. Kabluchko, J. Prochno, and C. Thäle. High-dimensional limit theorems for random vectors in ℓ_p^n-balls. II. Communications in Contemporary Mathematics, 23(3):1950073, 2021.
- [17]S. S. Kim and K. Ramanan. A conditional limit theorem for high-dimensional ℓ^p-spheres. Journal of Applied Probability, 55(4):1060–1077, 2018.
- [18]J. Prochno. The large and moderate deviations approach in geometric functional analysis. In R. Adamczak, N. Gozlan, K. Lounici, M. Madiman, F. Merlevède, and E. Werner, editors, High Dimensional Probability X: The Będlewo Volume, volume 82 of Progress in Probability, pages 347–421. Birkhäuser, Cham, 2026.arxiv.org/abs/2403.03940
- [19]S. T. Rachev and L. Rüschendorf. Approximate independence of distributions on spheres and their stability properties. The Annals of Probability, 19(3):1311–1337, 1991.DOI
- [20]G. Schechtman and M. Schmuckenschläger. Another remark on the volume of the intersection of two L^n_p balls. In J. Lindenstrauss and V. D. Milman, editors, Geometric Aspects of Functional Analysis: Israel Seminar (GAFA) 1989–90, volume 1469 of Lecture Notes in Mathematics, pages 174–178. Springer, Berlin, Heidelberg, 1991.DOI
- [21]G. Schechtman and J. Zinn. On the volume of the intersection of two L^n_p balls. Proceedings of the American Mathematical Society, 110(1):217–224, 1990.