arXiv subjects

Paata Ivanisvili

Publications and source records attributed to Paata Ivanisvili.

At least 37 records · Page 2Linked to original sources

arXiv2026

Almost-Orthogonality in Lp Spaces: A Case Study with Grok

Carbery proposed the following sharpened form of triangle inequality for many functions: for any $p\ge 2$ and any finite sequence $(f_j)_j\subset L^p$ we have \[ \Big\|\sum_j f_j\Big\|_p \ \le\ \left(\sup_{j} \sum_{k} α_{jk}^{\,c}\right)^{1/p'} \Big(\sum_j \|f_j\|_p^p\Big)^{1/p}, \] where $c=2$, $1/p+1/p'=1$, and $α_{jk}=\sqrt{\frac{\|f_{j}f_{k}\|_{p/2}}{\|f_{j}\|_{p}\|f_{k}\|_{p}}}$. In the first part of this paper we construct a counterexample showing that this inequality fails for every $p>2$. We then prove that if an estimate of the above form holds, the exponent must satisfy $c\le p'$. Finally, at the critical exponent $c=p'$, we establish the inequality for all integer values $p\ge 2$. In the second part of the paper we obtain a sharp three-function bound \[ \Big\|\sum_{j=1}^{3} f_j\Big\|_p \ \le\ \left(1+2Γ^{c(p)}\right)^{1/p'} \Big(\sum_{j=1}^{3} \|f_j\|_p^p\Big)^{1/p}, \] where $p \geq 3$, $c(p) = \frac{2\ln(2)}{(p-2)\ln(3)+2\ln(2)}$ and $Γ=Γ(f_1,f_2,f_3)\in[0,1]$ quantifies the degree of orthogonality among $f_1,f_2,f_3$. The exponent $c(p)$ is optimal, and improves upon the power $r(p) = \frac{6}{5p-4}$ obtained previously by Carlen, Frank, and Lieb. Some intermediate lemmas and inequalities appearing in this work were explored with the assistance of the large language model Grok.

math.CA↗

arXiv2026

Grokability in five inequalities

In this note, we report five mathematical discoveries made in collaboration with Grok, all of which have been subsequently verified by the authors. These include an improved lower bound on the maximal Gaussian perimeter of convex sets in $\mathbb{R}^n$, sharper $L_2$-$L_1$ moment comparison inequalities on the Hamming cube $\{-1,1\}^n$, a strengthened autoconvolution inequality, improved asymptotic bounds on the size of the largest $g$-Sidon sets in $\{1,\dots,n\}$, and an optimal balanced Szarek's inequality.

math.PR↗

arXiv2026

Sharp Lower Bounds for Dyadic Square Functions of indicator functions of sets

We study lower bounds for dyadic square functions of indicator functions. In the case of the dyadic square function $S_{2}$ we obtain a sharp lower bound: for every measurable $A \subset {[0,1)}$, we have \[ \|S_{2}(\mathbbm{1}_{A})\|_{1}\ge \mathbb{E}_{|A|}\big[\sqrtτ\big]\asymp |A|^{*}\log_2\frac{1}{|A|^{*}}, \] where $τ$ is the first exit time from $(0,1)$ of a standard Brownian motion started at $|A|$, and $|A|^{*}:=\min\{|A|,1-|A|\}$. This estimate gives logarithmic improvement over the classical Burkholder--Davis--Gundy lower bound $|A|^{*}$. In addition, we show a sharp inequality \[ \|S_{1}(\mathbbm{1}_{A})\|_{1} \ge T(|A|)\asymp |A|^{*}\log_{2}\frac{1}{|A|^{*}}, \] where $T(x)=\sum_{k=0}^{\infty}\frac{\operatorname{dist}(2^{k}x,\mathbb{Z})}{2^{k}}$ is the Takagi function.

math.CA↗

arXiv2026

Sharp isoperimetric inequalities on the Hamming cube II: The critical exponent

A sharp isoperimetric inequality for the Hamming cube is proved at the critical exponent $β=\frac12$. This follows up on previous work, where such bounds were established for $β$ near $\frac12$. As a consequence, this result settles a conjecture of Kahn and Park on cube partitions and yields a sharp $L^1$ Poincaré inequality for Boolean-valued functions. It also confirms a low-noise limit for balanced functions predicted by the Hellinger conjecture on noisy Boolean channels in information theory.

math.CA↗

arXiv2025

On the Eldan-Gross inequality

A recent discovery of Eldan and Gross states that there exists a universal $C>0$ such that for all Boolean functions $f:\{-1,1\}^n\to \{-1,1\}$, $$ \int_{\{-1,1\}^n}\sqrt{s_f(x)}dμ(x) \ge C\text{Var}(f)\sqrt{\log \left(1+\frac{1}{\sum_{j=1}^{n}\text{Inf}_j(f)^2}\right)} $$ where $s_f(x)$ is the sensitivity of $f$ at $x$, $\text{Var}(f)$ is the variance of $f$, $\text{Inf}_j(f)$ is the influence of $f$ along the $j$-th variable, and $μ$ is the uniform probability measure. In this note, we give an alternative proof that applies to biased discrete hypercube, and spaces having positive Ricci curvature lower bounds in the sense of Bakry and Émery.

math.FA↗

arXiv2025

Counterexample to majority optimality in NICD with erasures

We asked GPT-5 Pro to look for counterexamples among a public list of open problems (the Simons ``Real Analysis in Computer Science'' collection). After several numerical experiments, it suggested a counterexample for the Non-Interactive Correlation Distillation (NICD) with erasures question: namely, a Boolean function on 5 bits that achieves a strictly larger value of $\mathbb{E}|f(z)|$ than the 5-bit majority function when the erasure parameter is $p=0.40.$ In this very short note we record the finding, state the problem precisely, give the explicit function, and verify the computation step by step by hand so that it can be checked without a computer. In addition, we show that for each fixed odd $n$ the majority is optimal (among unbiased Boolean functions) in a neighborhood of $p=0$. We view this as a little spark of an AI contribution in Theoretical Computer Science: while modern Large Language Models (LLMs) often assist with literature and numerics, here a concrete finite counterexample emerged.

math.PR↗

arXiv2025

Covering the hypercube, the uncertainty principle, and an interpolation formula

We show that the minimal number of skewed hyperplanes that cover the hypercube $\{0,1\}^{n}$ is at least $\frac{n}{2}+1$, and there are infinitely many $n$'s when the hypercube can be covered with $n-\log_{2}(n)+1$ skewed hyperplanes. The minimal covering problems are closely related to uncertainty principle on the hypercube, where we also obtain an interpolation formula for multilinear polynomials on $\mathbb{R}^{n}$ of degree less than $\lfloor n/m \rfloor$ by showing that its coefficients corresponding to the largest monomials can be represented as a linear combination of values of the polynomial over the points $\{0,1\}^{n}$ whose hamming weights are divisible by $m$.

math.CO↗

arXiv2025

Optimal Young's convolutions inequality and its reverse form on the hypercube

We establish sharp forms of Young's convolution inequality and its reverse on the discrete hypercube $\{0,1\}^d$ in the diagonal case $p=q$. As applications, we derive bounds for additive energies and sumsets. We also investigate the non-diagonal regime $p\neq q$, providing necessary conditions for the inequality to hold, along with partial results in the case $r = 2$.

math.CA↗

arXiv2025

Multiversion of the Hausdorff--Young inequality

We consider a family of jointly Gaussian random vectors $ξ_j \in \mathbb{R}^{k_j}$, each standard normal but possibly correlated, and investigate when\[ \mathbb{E}\, F\!\Bigl(B\bigl(|T_{z_1} f_1(ξ_1)|,\dots,|T_{z_n} f_n(ξ_n)|\bigr)\Bigr) \;\;\le\;\; F\!\Bigl(\,\mathbb{E}\,B\bigl(|f_1(ξ_1)|,\dots,|f_n(ξ_n)|\bigr)\Bigr) \] holds, where $T_{z}$ is either a Mehler transform $(z \in \mathbb{C})$ or a noise operator $(z \in \mathbb{R})$. This framework unifies and extends real and complex hypercontractivity to multi-function settings, yielding multiversions of the sharp Hausdorff--Young inequality, the log-Sobolev inequality, and a noisy Gaussian--Jensen inequality. Applications include a new covariance-based characterization of the Brascamp--Lieb inequality in the presence of noise.

math.FA↗

arXiv2024

Jackson's inequality on the hypercube

We investigate the best constant $J(n,d)$ such that Jackson's inequality \[ \inf_{\mathrm{deg}(g) \leq d} \|f - g\|_{\infty} \leq J(n,d) \, s(f), \] holds for all functions $f$ on the hypercube $\{0,1\}^n$, where $s(f)$ denotes the sensitivity of $f$. We show that the quantity $J(n, 0.499n)$ is bounded below by an absolute positive constant, independent of $n$. This complements Wagner's theorem, which establishes that $J(n,d)\leq 1 $. As a first application we show that reverse Bernstein inequality fails in the tail space $L^{1}_{\geq 0.499n}$ improving over previously known counterexamples in $L^{1}_{\geq C \log \log (n)}$. As a second application, we show that there exists a function $f : \{0,1\}^n \to [-1,1]$ whose sensitivity $s(f)$ remains constant, independent of $n$, while the approximate degree grows linearly with $n$. This result implies that the sensitivity theorem $s(f) \geq Ω(\mathrm{deg}(f)^C)$ fails in the strongest sense for bounded real-valued functions even when $\mathrm{deg}(f)$ is relaxed to the approximate degree. We also show that in the regime $d = (1 - δ)n$, the bound \[ J(n,d) \leq C \min\{δ, \max\{δ^2, n^{-2/3}\}\} \] holds. Moreover, when restricted to symmetric real-valued functions, we obtain $J_{\mathrm{symmetric}}(n,d) \leq C/d$ and the decay $1/d$ is sharp. Finally, we present results for a subspace approximation problem: we show that there exists a subspace $E$ of dimension $2^{n-1}$ such that $\inf_{g \in E} \|f - g\|_{\infty} \leq s(f)/n$ holds for all $f$.

math.FA↗

arXiv2024

Low-degree learning and the metric entropy of polynomials

Let $\mathscr{F}_{n,d}$ be the class of all functions $f:\{-1,1\}^n\to[-1,1]$ on the $n$-dimensional discrete hypercube of degree at most $d$. In the first part of this paper, we prove that any (deterministic or randomized) algorithm which learns $\mathscr{F}_{n,d}$ with $L_2$-accuracy $\varepsilon$ requires at least $Ω((1-\sqrt{\varepsilon})2^d\log n)$ queries for large enough $n$, thus establishing the sharpness as $n\to\infty$ of a recent upper bound of Eskenazis and Ivanisvili (2021). To do this, we show that the $L_2$-packing numbers $\mathsf{M}(\mathscr{F}_{n,d},\|\cdot\|_{L_2},\varepsilon)$ of the concept class $\mathscr{F}_{n,d}$ satisfy the two-sided estimate $$c(1-\varepsilon)2^d\log n \leq \log \mathsf{M}(\mathscr{F}_{n,d},\|\cdot\|_{L_2},\varepsilon) \leq \frac{2^{Cd}\log n}{\varepsilon^4}$$ for large enough $n$, where $c, C>0$ are universal constants. In the second part of the paper, we present a logarithmic upper bound for the randomized query complexity of classes of bounded approximate polynomials whose Fourier spectra are concentrated on few subsets. As an application, we prove new estimates for the number of random queries required to learn approximate juntas of a given degree, functions with rapidly decaying Fourier tails and constant depth circuits of given size. Finally, we obtain bounds for the number of queries required to learn the polynomial class $\mathscr{F}_{n,d}$ without error in the query and random example models.

cs.LG↗

arXiv2024

$(P,Q)$ complex hypercontractivity

Let $ξ$ be the standard normal random vector in $\mathbb{R}^{k}$. Under some mild growth and smoothness assumptions on any increasing $P, Q : [0, \infty) \mapsto [0, \infty)$ we show $(P,Q)$ complex hypercontractivity $$ Q^{-1}(\mathbb{E} Q(|T_{z}f(ξ)|))\leq P^{-1}(\mathbb{E}P(|f(ξ)|)) $$ holds for all polynomials $f:\mathbb{R}^{k} \mapsto \mathbb{C}$, where $T_{z}$ is the hermite semigroup at complex parameter $z, |z|\leq 1$, if and only if \begin{align*} \left|\frac{tP''(t)}{P'(t)}-z^{2}\frac{tQ''(t)}{Q'(t)}+z^{2}-1\right|\leq \frac{tP''(t)}{P'(t)}-|z|^{2}\frac{tQ''(t)}{Q'(t)}+1-|z|^{2} \end{align*} holds for all $t>0$ provided that $F''>0$, and $F'/F''$ is concave, where $F = Q\circ P^{-1}$. This extends Hariya's result from real to complex parameter $z$. Several old and new applications are presented for different choices of $P$ and $Q$. The proof uses heat semigroup arguments, where we find a certain map $C(s)$, which interpolates the inequality at the endpoints. The map $C(s)$ itself is composed of four heat flows running together at different times.

math.FA↗

arXiv2024

Sharpening the gap between $L^{1}$ and $L^{2}$ norms

We refine the classical Cauchy--Schwartz inequality $\|X\|_{1} \leq \|X\|_{2}$ by demonstrating that for any $p$ and $q$ with $q>p>2$, there exists a constant $C=C(p,q)$ such that $\|X\|_1 \leq 1 - C \Big{(}\|X\|_p^p - 1\Big{)}^{\frac{q-2}{q-p}}\Big{(}\|X\|_q^q - 1\Big{)}^{\frac{2-p}{q-p}}$ holds true for all Borel measurable random variables $X$ with $\|X\|_{2}=1$ and $\|X\|_{p}<\infty$. We illustrate two applications of this result: one for biased Rademacher sums and another for exponential sums.

math.PR↗

arXiv2024

The KKL inequality and Rademacher type 2

We show that a vector-valued Kahn--Kalai--Linial inequality holds in every Banach space of Rademacher type 2. We also show that for any nondecreasing function $h\geq 0$ with $0<\int_{1}^{\infty}\frac{h(t)}{t^{2}}\mathrm{dt}<\infty$ we have the inequality \begin{align*} \|f - \mathbb{E}f\|_2 \leq 12 \, T_{2}(X) \left(\int_{1}^{\infty}\frac{h(t)}{t^{2}} \mathrm{dt} \right)^{1/2} \, \left(\sum_{j=1}^n \frac{\|D_j f\|^{2}_2}{h\left( \log \frac{\|D_j f\|_2}{\|D_j f\|_1} \right)}\right)^{1/2} \end{align*} for all $f :\{-1,1\}^{n} \to X$ and all $n\geq 1$, where $X$ is a normed space and $T_{2}(X)$ is the associated type 2 constant.

math.PR↗

arXiv2024

Discrete Brunn-Minkowski Inequality for subsets of the cube

We show that for all $A, B \subseteq \{0,1,2\}^{d}$ we have $$ |A+B|\geq (|A||B|)^{\log(5)/(2\log(3))}. $$ We also show that for all finite $A,B \subset \mathbb{Z}^{d}$, and any $V \subseteq\{0,1\}^{d}$ the inequality $$ |A+B+V|\geq |A|^{1/p}|B|^{1/q}|V|^{\log_{2}(p^{1/p}q^{1/q})} $$ holds for all $p \in (1, \infty)$, where $q=\frac{p}{p-1}$ is the conjugate exponent of $p$. All the estimates are dimension free with the best possible exponents. We discuss applications to various related problems.

math.CO↗

arXiv2023

Additive energies on discrete cubes

We prove that for $d\geq 0$ and $k\geq 2$, for any subset $A$ of a discrete cube $\{0,1\}^d$, the $k-$higher energy of $A$ (the number of $2k-$tuples $(a_1,a_2,\dots,a_{2k})$ in $A^{2k}$ with $a_1-a_2=a_3-a_4=\dots=a_{2k-1}-a_{2k}$) is at most $|A|^{\log_{2}(2^k+2)}$, and $\log_{2}(2^k+2)$ is the best possible exponent. We also show that if $d\geq 0$ and $2\leq k\leq 10$, for any subset $A$ of a discrete cube $\{0,1\}^d$, the $k-$additive energy of $A$ (the number of $2k-$tuples $(a_1,a_2,\dots,a_{2k})$ in $A^{2k}$ with $a_1+a_2+\dots+a_k=a_{k+1}+a_{k+2}+\dots+a_{2k}$) is at most $|A|^{\log_2{ \binom{2k}{k}}}$, and $\log_2{ \binom{2k}{k}}$ is the best possible exponent. We discuss the analogous problems for the sets $\{0,1,\dots,n\}^d$ for $n\geq 2$.

math.CO↗

arXiv2023

Bellman functions on simple non-convex domains in the plane

The present paper provides a generalization of the previous authors' work on Bellman functions for integral functionals on $\mathrm{BMO}$. Those Bellman functions are the minimal locally concave functions on parabolic strips in the plane. Now we describe the algorithm for constructing minimal locally concave functions on a planar domain that is a difference of two unbounded convex domains. This leads to many sharp estimates for functions in the classes like $\mathrm{BMO}$, $A_p$, or the Gehring classes.

math.CA↗

arXiv2023

On sharp isoperimetric inequalities on the hypercube

We prove the sharp isoperimetric inequality $$ \mathbb{E} \,h_{A}^{\log_{2}(3/2)} \geq μ(A)^{*} (\log_{2}(1/μ(A)^{*}))^{\log_{2}(3/2)} $$ for all sets $A \subseteq \{0,1\}^n$, where $μ$ denotes the uniform probability measure, $μ(A)^{*}=\min\{μ(A), 1-μ(A)\}$, $h_A$ is supported on $A$ and to each vertex $x$ assigns the number of neighbour vertices in the complement of $A$. The inequality becomes equality for any subcube. Moreover, we provide lower bounds on $\mathbb{E} h_{A}^β$ in terms of $μ(A)$ for all $β\in [1/2,1]$, improving, and in some cases tightening, previously known results. In particular, we obtain the sharp inequality $\mathbb{E}h_{A}^{0.53}\geq 2 μ(A)(1-μ(A))$ for all sets with $μ(A)\geq 1/2$, which allows us to refine a recent result of Kahn and Park on isoperimetric inequalities about partitioning the hypercube. Furthermore, we derive Talagrand's isoperimetric inequalities for functions with values in a Banach space having finite cotype: for all $f :\{-1,1\}^{n} \to X$, $\|f\|_{\infty}\leq 1$, and any $p \in [1,2]$ we have $$ \|Df\|_{p} \gtrsim \frac{1}{q^{3/2}C_{q}(X)} \|f\|_{2}^{2/p}\left(\log \frac{e\|f\|_{2}}{\|f\|_{1}}\right)^{1/q}, $$ where $\| Df\|_{p}^{p} = \mathbb{E} \| \sum_{1\leq j \leq n} x'_{j} D_{j} f(x)\|^{p}$, $x'$ is independent copy of $x$, and $C_{q}(X)$ is the cotype $q$ constant of $X$. Different proofs of the recently resolved Talagrand's conjecture will be presented.

math.CA↗