Distribution of random multiplicative functions in short intervals, with proper normalization

Adam J Harper, Kannan Soundararajan, Max Wenqiang Xu · 2026

Read Distribution of random multiplicative functions in short intervals, with proper normalization by Adam J Harper, Kannan Soundararajan, Max Wenqiang Xu (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2606-29040.txt.

Introduction

Harper, Soundararajan and Xu determine how sums of a random multiplicative function behave over short intervals of integers. A Steinhaus random multiplicative function assigns random values to primes and multiplies them across each number's factorisation, mimicking arithmetic functions. They study partial sums over intervals that grow, but stay small compared with their starting point.

With the right normalisation, they prove the limiting distribution of these sums is Gaussian, the bell curve, for all such interval lengths. A key new feature is the normalisation itself: when the interval length is very close to the starting point, the correct divisor differs from the standard deviation one would naively use. In contrast, when the interval is comparable in size to its start, they show no normalisation can produce a non-degenerate Gaussian limit. The result draws a precise line between regimes where bell-curve behaviour holds and where it cannot.

Catalog record

CREATORAdam J Harper, Kannan Soundararajan, Max Wenqiang Xu
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE151 KB
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2606.29040

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Distribution of random multiplicative functions in short intervals, with proper normalization Adam J Harper, Kannan Soundararajan, Max Wenqiang Xu Abstract . We determine the limiting distribution of partial sums of a Steinhaus random multiplicative function ∑ x ≤ n ≤ x + y f ⁡ ( n ) \sum_{x\leq n\leq x+y}f(n) over short intervals [ x , x + y ] [x,x+y] , where y → ∞ y\rightarrow\infty but y = o ⁡ ( x ) y=o(x) . We show that with appropriate normalization, the limiting distribution is Gaussian for all such y y . A key new feature of our result is that the normalization factor is different from the standard deviation y \sqrt{y} when y y is very close to x x . In contrast, when y ≍ x y\asymp x there is no normalization for which the limiting distribution is a non-degenerate Gaussian. 1. Introduction The study of random multiplicative functions has been very active in recent years, motivated by various connections and applications in number theory, probability, and analysis. Define a Steinhaus random multiplicative function f:\mathbb{N}\rightarrow\mathbb{C} , by taking (f(p))_{p\;\text{prime}} to be independent random variables distributed uniformly on the complex unit circle, and setting f(n):=\prod_{p^{a}\|n}f(p)^{a} for all natural numbers n (where p^{a}\|n means that p^{a} is the highest power of the prime p that divides n ). These are often considered to be models for number-theoretic functions like Dirichlet characters \chi(n) . A Rademacher random multiplicative function is defined by letting (f(p))_{p\;\text{prime}} be independent, taking values \pm 1 with probability \frac{1}{2} each, and setting f(n):=\prod_{p|n}f(p) for all squarefree n , and f(n)=0 when n is not squarefree. A fundamental question in the area is to study the statistical properties of partial sums \sum_{1\leq n\leq x}f(n) , and more generally, weighted partial sums \sum_{1\leq n\leq x}a(n)f(n) . One striking result is Harper’s resolution [ 14 ] of Helson’s conjecture [ 16 ] . Harper proved (in both the Steinhaus and Rademacher settings) that for all large x (1.1) \mathbb{E}\Big[\Big|\sum_{1\leq n\leq x}f(n)\Big|\Big]\asymp\frac{\sqrt{x}}{(\log\log x)^{\frac{1}{4}}}, so that partial sums of random multiplicative functions exhibit “better than square-root cancellation.” This implies that the normalized partial sum, with the natural normalization \Big(\mathbb{E}\Big[\Big|\sum_{1\leq n\leq x}f(n)\Big|^{2}\Big]\Big)^{-\frac{1}{2}}\asymp\frac{1}{\sqrt{x}}, has a trivial limiting distribution. In earlier work [ 12 ] , it had been proved (in the Rademacher case) that \Big(\mathbb{E}\Big[\Big|\sum_{1\leq n\leq x}f(n)\Big|^{2}\Big]\Big)^{-\frac{1}{2}}\sum_{n\leq x}f(n) could not converge in distribution to a standard Gaussian N(0,1) (though this did not exclude the more subtle possibility that it might converge to some Gaussian with variance different from 1 ). A central limit theorem may emerge in cases with general weights a(n) that somewhat disrupt the multiplicative structure. One particular example that has attracted considerable attention is when a(n) is an indicator function of a short interval, namely we study the limiting distribution of \sum_{x\leq n\leq x+y}f(n) where y\rightarrow\infty , but y=o(x) . It was first proved by Chatterjee and Soundararajan in [ 3 ] (for Rademacher f(n) ) that the limiting distribution, with usual normalization 1/\sqrt{y} , is Gaussian as long as Cx^{1/5}\log x\leq y=o(x/\log x) . Note that the lower bound condition on y is, in a sense, much less interesting here; it was imposed so that one knows unconditionally that all intervals [x,x+y] contain roughly the expected quantity of squarefree numbers. This result was improved by Soundararajan and Xu in [ 21 ] to the wider range Cx^{1/5}\log x\leq y\leq x/(\log x)^{\log 4-1+\epsilon} , and also extended to allow Steinhaus f(n) whenever y\rightarrow\infty and y\leq x/(\log x)^{\log 4-1+\epsilon} . On the other hand, by using the triangle inequality and Harper’s theorem ( 1.1 ), one can see that when y is sufficiently close to x , \frac{1}{\sqrt{y}}\sum_{x\leq n\leq x+y}f(n) has a trivial limiting distribution. In a recent work of Caich [ 1 ] , it is proved that such trivial limiting distribution holds as long as y\geq\frac{x}{\exp((\log\log x)^{1/2-\epsilon})} . It has been an intriguing question to properly understand all these transitions in behavior. In particular, one would like to know whether \frac{1}{\sqrt{y}}\sum_{x\leq n\leq x+y}f(n) actually has Gaussian behavior for y larger than x/(\log x)^{\log 4-1+\epsilon} , and what can be said about the distribution for those very large y where the limit becomes trivial. Perhaps surprisingly, we prove a Gaussian limit theorem for partial sums \sum_{x\leq n\leq x+y}f(n) of a Steinhaus random multiplicative function whenever y\rightarrow\infty with y=o(x) . This completely determines the distribution over all short intervals, thereby completing the previous results in [ 3 , 21 , 1 ] . Theorem 1.1 . Let f(n) be a Steinhaus random multiplicative function. There exists a deterministic scaling factor V(x,y) such that the quantity \frac{1}{\sqrt{V(x,y)}}\sum_{x\leq n\leq x+y}f(n) converges in distribution to a standard complex Gaussian random variable with mean 0 and variance 1 , as y\to\infty with y=o(x) . The scaling factor satisfies (1.2) V(x,y)\sim\frac{y}{\sqrt{2\pi}}\int_{-\kappa}^{\kappa}e^{-u^{2}/2}du,\qquad\text{with}\qquad\kappa=\frac{\log(x/y)}{\sqrt{2\log\log x}}, with the asymptotic holding as x/y\to\infty . In particular (1.3) V(x,y)\asymp y\cdot\min\Big\{1,\frac{\log(x/y)}{\sqrt{\log\log x}}\Big\}, and V(x,y)\sim y as \frac{\log(x/y)}{\sqrt{\log\log x}}\to+\infty . The value of V(x,y) that emerges naturally from our arguments is expressed in terms of a quantity \mu(x,y) , which is (up to scaling) a certain random walk probability. See Proposition 3.5 and ( 4.3 ) below, and the discussion in sections 1.1 and 1.2 . However, we are free to replace V(x,y) in Theorem 1.1 by any quantity to which it is asymptotic (as y\to\infty with y=o(x) ), and our estimation of \mu(x,y) (again see Proposition 3.5 ) shows that ( 1.2 ) is such a choice. This is the first time that partial sums of random multiplicative functions with non-obvious normalization have been found to have a Gaussian limiting distribution. There are a few other number-theoretic situations where somewhat similar (and very interesting) phenomena arise. For example, Montgomery and Soundararajan [ 20 ] show, assuming various strong but plausible hypotheses, that the distribution of the von Mangoldt function in short intervals \sum_{x\leq n\leq x+y}\Lambda(n) as x\leq N varies and with y in the range N^{\delta}\leq y\leq N^{1-\delta} , should be roughly Gaussian, but with variance differing from the obvious guess by a constant factor. Gorodetsky, Mangerel and Rodgers [ 4 ] recently gave a beautiful proof that the analogous count of squarefree numbers in short intervals, where x\leq N varies and y\rightarrow\infty with y=N^{o(1)} , is roughly Gaussian with variance differing from the obvious guess by a power. In the squarefree case, the variance drop reflects the highly structured nature of the sequence of squarefree numbers. For the von Mangoldt function (equivalently primes) the situation is closer to our case here, but again the variance drop reflects a structural property of the sequence, namely its connection with the zeros of the Riemann zeta function. We are not aware of a similar phenomenon having been observed before in any random or deterministic context quite analogous to the present one (e.g. involving character sums). We also note that in both the primes and the squarefree examples, one sees the variance drop simply by computing the variance of the sequence itself (and the results on limiting distributions are obtained by a moment method). In contrast, the variance of our random sums has the obvious size one would expect (from orthogonality), it is only in the limiting distribution that one sees a change of

Read the complete text