Papers by Luka Milićević
22 paper(s) by this author
· All BibTeX
A note on the shortest law for the symmetric group
Let $α(n)$ denote the length of the shortest non-trivial two-variable law for the symmetric group $S_n$. Buskin's quantitative subgroup-separability argument gives the classical lower bound $α(n)\geq 2n-O(1)$. In this short note we give an improvement by proving that $α(n)\geq \frac52 n-O(1)$.
General inverse theory for the $\mathsf{U}^4$ norm
In this paper, we develop a quantitative inverse theory for the Gowers uniformity norm $\|\cdot\|_{\mathsf{U}^4}$ in general finite abelian groups. We identify a new type of obstructions to uniformity, which we call almost-cubic polynomials. An almost-cubic polynomial $q$ on a Bohr set $B(Γ, ρ_0)$ is a function such that, for each $ρ\leq \min\{ρ_0, 1/8\}$, we have \[\|Δ_{a,b,c,d} q(x)\|_{\mathbb{T}} \leq 2^{10} ρ\] for all $x, a,b,c,d \in B(Γ, ρ)$.
Let $f : G \to \mathbb{D}$ be a function with $\|f\|_{\mathsf{U}^4} \geq c$. We prove quasipolynomial inverse theorems:
$\bullet$ when $(|G|, 6) = 1$, there exists an almost-cubic $q : B(Γ, ρ)$ for $|Γ| \leq \log^{O(1)} c^{-1}$ and $ρ\geq \exp(-\log^{O(1)} c^{-1})$, and an element $t \in G$ such that $$\Big|\sum_{x \in G} 1_{B}(x) f(x + t) \operatorname{e}(q(x))\Big| \geq \exp(-\log^{O(1)} c^{-1})|G|,$$ $\bullet$ when $G = (\mathbb{Z}/2^d\mathbb{Z})^n$, there exists a cubic polynomial $q : G \to \mathbb{T}$ such that $$\Big|\sum_{x \in G} f(x)\operatorname{e}(q(x))\Big| \geq \exp(-\log^{O_d(1)} c^{-1})|G|.$$ Almost-cubic polynomials are rather rigid and we exhibit a strong connection with generalized polynomials in the case of cyclic groups, as well as with polynomials in the classical sense in the case of finite vector spaces. We also answer a question of Jamneshan, Shalom and Tao concerning the inverse theory in groups of bounded torsion.
The central result from which the inverse theorems follow is a structural result for Freiman bihomomorphisms in general finite abelian groups. In our proof, we generalize methods of our previous work in the case of finite vector spaces and introduce novel ideas concerning extensions of Freiman bihomomorphisms. In the problem of extension of Freiman bihomomorphisms, genuinely new phenomena appear in general finite abelian groups.
Möbius function is strongly orthogonal to polynomial phases over $\mathbb{F}_p[t]$
In this paper, we prove power-saving bounds for the corelation of the Möbius function with polynomial phases of degree $k$ in function fields $\mathbb{F}_p[t]$, when $p > k$. The proof relies on a new approximation result for phases of biased multilinear forms and the recently established strong bounds for the problem of finding bounded codimension varieties inside the dense ones. Along the way, we also obtain polynomial bounds in the inverse theorem for Gowers uniformity norms in the special case of polynomial phases in finite vector spaces.
Low-codimensional Subvarieties Inside Dense Multilinear Varieties
Published in Proceedings of the 13th European Conference on Combinatorics, Graph Theory and Applications EUROCOMB '25 (2025), 904--909
• Search Publication
Let $G_1, \dots, G_k$ be finite-dimensional vector spaces over a prime field $\mathbb{F}_p$. Let $V$ be a variety inside $G_1 \times \cdots \times G_k$ defined by a multilinear map. We show that if $|V| \geq c |G_1| \cdots |G_k|$, then $V$ contains a subvariety defined by at most $K(\log_{p} c^{-1} + 1)$ multilinear forms, where $K$ depends on $k$ only. This result is optimal up to multiplicative constant and is relevant to the partition vs. analytic rank problem in additive combinatorics.
Quasipolynomial inverse theorem for the $\mathsf{U}^4(\mathbb{F}_p^n)$ norm
The inverse theory for Gowers uniformity norms is one of the central topics in additive combinatorics and one of the most important aspects of the theory is the question of bounds. In this paper, we prove a quasipolynomial inverse theorem for the $\mathsf{U}^4$ norm in finite vector spaces. The proof follows a different strategy compared to the existing quantitative inverse theorems. In particular, the argument relies on a novel argument, which we call the abstract Balog-Szemerédi-Gowers theorem, and combines several other ingredients such as algebraic regularity method, bilinear Bogolyubov argument and algebraic dependent random choice.
Good bounds for sets lacking skew corners
A skew corner is a triple of points in $\mathbb{Z} \times \mathbb{Z}$ of the form $(x,y), (x, y + a)$ and $(x + a, y')$. Pratt posed the following question: how large can a set $A \subseteq [n] \times [n]$ be, provided it contains no non-trivial skew corner (i.e. one for which $a\not=0$)? We prove that $|A| \leq \exp(- c\log^c n) n^2$, for an absolute constant $c > 0$, which, along with a construction of Beker, essentially resolves Pratt's question.
Our argument is represents a two-dimensional variant of the method of Kelley and Meka, which they used to prove Behrend-type bounds in Roth's theorem. A very similar result was obtained independently and simultaneously by Jaber, Lovett and Ostuni.
A collection of open problems in celebration of Imre Leader's 60th birthday
One of the great pleasures of working with Imre Leader is to experience his infectious delight on encountering a compelling combinatorial problem. This collection of open problems in combinatorics has been put together by a subset of his former PhD students and students-of-students for the occasion of his 60th birthday. All of the contributors have been influenced (directly or indirectly) by Imre: his personality, enthusiasm and his approach to mathematics. The problems included cover many of the areas of combinatorial mathematics that Imre is most associated with: including extremal problems on graphs, set systems and permutations, and Ramsey theory. This is a personal selection of problems which we find intriguing and deserving of being better known. It is not intended to be systematic, or to consist of the most significant or difficult questions in any area. Rather, our main aim is to celebrate Imre and his mathematics and to hope that these problems will make him smile. We also hope this collection will be a useful resource for researchers in combinatorics and will stimulate some enjoyable collaborations and beautiful mathematics.
A note on transverse sets and bilinear varieties
Published
• View Publication
• BIB
Let $G$ and $H$ be finite-dimensional vector spaces over $\mathbb{F}_p$. A subset $A \subseteq G \times H$ is said to be transverse if all of its rows $\{x \in G \colon (x,y) \in A\}$, $y \in H$, are subspaces of $G$ and all of its columns $\{y \in H \colon (x,y) \in A\}$, $x \in G$, are subspaces of $H$. As a corollary of a bilinear version of Bogolyubov argument, Gowers and the author proved that dense transverse sets contain bilinear varieties of bounded codimension. In this paper, we provide a direct combinatorial proof of this fact. In particular, we improve the bounds and evade the use of Fourier analysis and Freiman's theorem and its variants.
Approximate quadratic varieties
A classical result in additive combinatorics, which is a combination of Balog-Szemerédi-Gowers theorem and a variant of Freiman's theorem due to Ruzsa, says that if a subset $A$ of $\mathbb{F}_p^n$ contains at least $c |A|^3$ additive quadruples, then there exists a subspace $V$, comparable in size to $A$, such that $|A \cap V| \geq Ω_c(|A|)$. Motivated by the fact that higher order approximate algebraic structures play an important role in the theory of uniformity norms, it would be of interest to find higher order analogues of the mentioned result.
In this paper, we study a quadratic version of the approximate property in question, namely what it means for a set to be an approximate quadratic variety. It turns out that information on the number of additive cubes, which are 8-tuples of the form $(x, x+ a,$ $x+ b, x+ c,$ $x+ a + b, x+ a + c,$ $x+ b + c, x+ a + b + c)$, in a set is insufficient on its own to guarantee quadratic structure, and it is necessary to restrict linear structure in a given set, which is a natural assumption in this context. With this in mind, we say that a subset $V$ of a finite vector space $G$ is a $(c_0, δ, \varepsilon)$-approximate quadratic variety if $|V| = δ|G|$, $\|1_V - δ\|_{\mathsf{U}^2} \leq \varepsilon$ and $V$ contains at least $c_0δ^7 |G|^4$ additive cubes. Our main result is the structure theorem for approximate quadratic varieties, stating that such a set has a large intersection with an exact quadratic variety of comparable size.
Quantitative inverse theorem for Gowers uniformity norms $\mathsf{U}^5$ and $\mathsf{U}^6$ in $\mathbb{F}_2^n$
We prove quantitative bounds for the inverse theorem for Gowers uniformity norms $\mathsf{U}^5$ and $\mathsf{U}^6$ in $\mathbb{F}_2^n$. The proof starts from an earlier partial result of Gowers and the author which reduces the inverse problem to a study of algebraic properties of certain multilinear forms. The bulk of the work in this paper is a study of the relationship between the natural actions of $\operatorname{Sym}_4$ and $\operatorname{Sym}_5$ on the space of multilinear forms and the partition rank, using an algebraic version of regularity method. Along the way, we give a positive answer to a conjecture of Tidor about approximately symmetric multilinear forms in 5 variables, which is known to be false in the case of 4 variables. Finally, we discuss the possible generalization of the argument for $\mathsf{U}^k$ norms.
Approximately Symmetric Forms Far From Being Exactly Symmetric
Let $V$ be a finite-dimensional vector space over $\mathbb{F}_p$. We say that a multilinear form $α\colon V^k \to \mathbb{F}_p$ in $k$ variables is $d$-approximately symmetric if the partition rank of difference $α(x_1, \dots, x_k) - α(x_{π(1)}, \dots, x_{π(k)})$ is at most $d$ for every permutation $π\in \operatorname{Sym}_k$. In a work concerning the inverse theorem for the Gowers uniformity $\|\cdot\|_{\mathsf{U}^4}$ norm in the case of low characteristic, Tidor conjectured that any $d$-approximately symmetric multilinear form $α\colon V^k \to \mathbb{F}_p$ differs from a symmetric multilinear form by a multilinear form of partition rank at most $O_{p,k,d}(1)$ and proved this conjecture in the case of trilinear forms. In this paper, somewhat surprisingly, we show that this conjecture is false. In fact, we show that approximately symmetric forms can be quite far from the symmetric ones, by constructing a multilinear form $α\colon \mathbb{F}_2^n \times \mathbb{F}_2^n \times \mathbb{F}_2^n \times \mathbb{F}_2^n \to \mathbb{F}_2$ which is 3-approximately symmetric, while the difference between $α$ and any symmetric multilinear form is of partition rank at least $Ω(\sqrt[3]{n})$.
An Inverse Theorem for Certain Directional Gowers Uniformity Norms
Published
• View Publication
• BIB
Let $G$ be a finite-dimensional vector space over a prime field $\mathbb{F}_p$ with some subspaces $H_1, \dots, H_k$. Let $f \colon G \to \mathbb{C}$ be a function. Generalizing the notion of Gowers uniformity norms, Austin introduced directional Gowers uniformity norms of $f$ over $(H_1, \dots, H_k)$ as \[\|f\|_{\mathsf{U}(H_1, \dots, H_k)}^{2^k} = \mathbb{E}_{x \in G,h_1 \in H_1, \dots, h_k \in H_k} \partial_{h_1} \dots \partial_{h_k} f(x)\] where $\partial_u f(x) \colon= f(x + u) \overline{f(x)}$ is the discrete multiplicative derivative.
Suppose that $G$ is a direct sum of subspaces $G = U_1 \oplus U_2 \oplus \dots \oplus U_k$. In this paper we prove the inverse theorem for the norm \[\|\cdot\|_{\mathsf{U}(U_1, \dots, U_k, \smash[b]{\underbrace{\scriptstyle G, \dots, G}_{\scriptscriptstyle \ell}})},\] which is the simplest interesting unknown case of the inverse problem for the directional Gowers uniformity norms. Namely, writing $\|\cdot\|_{\mathsf{U}}$ for the norm above, we show that if $f \colon G \to \mathbb{C}$ is a function bounded by 1 in magnitude and obeying $\|f\|_{\mathsf{U}} \geq c$, provided $\ell < p$, one can find a polynomial $α\colon G \to \mathbb{F}_p$ of degree at most $k + \ell - 1$ and functions $g_i \colon \oplus_{j \in [k] \setminus \{i\}} G_j \to \{z \in \mathbb{C} \colon |z| \leq 1\}$ for $i \in [k]$ such that \[\Big|\mathbb{E}_{x \in G} f(x) ω^{α(x)} \prod_{i \in [k]} g_i(x_1, \dots, x_{i-1}, x_{i+1}, \dots, x_k)\Big| \geq \Big(\exp^{(O_{p,k,\ell}(1))}(O_{p,k,\ell}(c^{-1}))\Big)^{-1}.\]
The proof relies on an approximation theorem for the cuboid-counting function that is proved using the inverse theorem for Freiman multi-homomorphisms.
Polynomial bound for partition rank in terms of analytic rank
Let $G_1, \dots, G_k$ be vector spaces over a finite field $\mathbb{F} = \mathbb{F}_q$ with a non-trivial additive character $χ$. The analytic rank of a multilinear form $α\colon G_1 \times \dots \times G_k \to \mathbb{F}$ is defined as $\operatorname{arank}(α) = -\log_q \mathbb{E}_{x_1 \in G_1, \dots, x_k\in G_k} χ\big(α(x_1,\dots, x_k)\big)$. The partition rank $\operatorname{prank}(α)$ of $α$ is the smallest number of maps of partition rank 1 that add up to $α$, where a map is of partition rank 1 if it can be written as a product of two multilinear forms, depending on different coordinates. It is easy to see that $\operatorname{arank}(α) \leq O\Big(\operatorname{prank}(α)\Big)$ and it has been known that $\operatorname{prank}(α)$ can be bounded from above in terms of $\operatorname{arank}(α)$. In this paper, we improve the latter bound to polynomial, i.e. we show that there are quantities $C, D$ depending on $k$ only such that $\operatorname{prank}(α) \leq C (\operatorname{arank}(α)^D + 1)$. As a consequence, we prove a conjecture of Kazhdan and Ziegler. The same result was obtained independently and simultaneously by Janzer.
An improved upper bound for the grid Ramsey problem
For a positive integer $r$, let $G(r)$ be the smallest $N$ such that, whenever the edges of the Cartesian product $K_N \times K_N$ are $r$-coloured, then there is a rectangle in which both pairs of opposite edges receive the same colour. In this paper, we improve the upper bounds on $G(r)$ by proving $G(r) \leq \Big(1 - \frac{1}{128}r^{-2}\Big) r^{\binom{r+1}{2}}$, for $r$ large enough. Unlike the previous improvements, which were based on bounds for the size of set systems with restricted intersection sizes, our proof is a form of a quasirandomness argument.
Classification theorem for strong triangle blocking arrangements
Published
• View Publication
• BIB
A strong triangle blocking arrangement is a geometric arrangement of some line segments in a triangle with certain intersection properties. It turns out that they are closely related to blocking sets. Our aim in this paper is to prove a classification theorem for strong triangle blocking arrangements. As an application, we obtain a new proof of the result of Ackerman, Buchin, Knauer, Pinchasi and Rote which says that $n$ points in general position cannot be blocked by $n-1$ points, unless $n = 2,4$. We also conjecture an extremal variant of the blocking points problem.
A quantitative inverse theorem for the $U^4$ norm over finite fields
A remarkable result of Bergelson, Tao and Ziegler implies that if $c>0$, $k$ is a positive integer, $p\geq k$ is a prime, $n$ is sufficiently large, and $f:\mathbb F_p^n\to\mathbb C$ is a function with $\|f\|_\infty\leq 1$ and $\|f\|_{U^k}\geq c$, then there is a polynomial $π$ of degree at most $k-1$ such that $\mathbb E_xf(x)ω^{-π(x)}\geq c'$, where $ω=\exp(2πi/p)$ and $c'>0$ is a constant that depends on $c,k$ and $p$ only. A version of this result for low-characteristic was also proved by Tao and Ziegler. The proofs of these results do not yield a lower bound for $c'$. Here we give a different proof in the high-characteristic case when $k=4$, which enables us to give an explicit estimate for $c'$. The bound we obtain is roughly doubly exponential in the other parameters.
Covering complete graphs by monochromatically bounded sets
Published
• View Publication
• BIB
Given a $k$-colouring of the edges of the complete graph $K_n$, are there $k-1$ monochromatic components that cover its vertices? This important special case of the well-known Lovász-Ryser conjecture is still open. In this paper we consider a strengthening of this question, where we insist that the covering sets are not merely connected but have bounded diameter. In particular, we prove that for any colouring of $E(K_n)$ with 4 colours, there is a choice of sets $A_1, A_2, A_3$ that cover all vertices, and colours $c_1, c_2, c_3$, such that for each $i = 1,2,3$ the monochromatic subgraph induced by the set $A_i$ and the colour $c_i$ has diameter at most 160.
Small Sets with Large Difference Sets
For every $ε> 0$ and $k \in \mathbb{N}$, Haight constructed a set $A \subset \mathbb{Z}_N$ ($\mathbb{Z}_N$ stands for the integers modulo $N$) for a suitable $N$, such that $A-A = \mathbb{Z}_N$ and $|kA| < εN$. Recently, Nathanson posed the problem of constructing sets $A \subset \mathbb{Z}_N$ for given polynomials $p$ and $q$, such that $p(A) = \mathbb{Z}_N$ and $|q(A)| < εN$, where $p(A)$ is the set $\{p(a_1, a_2, \dots, a_n)\phantom{.}\colon\phantom{.}a_1, a_2, \dots, a_n \in A\}$, when $p$ has $n$ variables. In this paper, we give a partial answer to Nathanson's question. For every $k \in \mathbb{N}$ and $ε> 0$, we find a set $A \subset \mathbb{Z}_N$ for suitable $N$, such that $A- A = \mathbb{Z}_N$, but $|A^2 + kA| < εN$, where $A^2 + kA = \{a_1a_2 + b_1 + b_2 + \dots + b_k\phantom{.}\colon\phantom{.}a_1, a_2,b_1, \dots, b_k \in A\}$. We also extend this result to construct, for every $k \in \mathbb{N}$ and $ε> 0$, a set $A \subset \mathbb{Z}_N$ for suitable $N$, such that $A- A = \mathbb{Z}_N$, but $|3A^2 + kA| < εN$, where $3A^2 + kA = \{a_1a_2 + a_3a_4 + a_5a_6 + b_1 + b_2 + \dots + b_k\phantom{.}\colon\phantom{.}a_1, \dots, a_6,b_1, \dots, b_k \in A\}$.
Decomposing the Complete $r$-Graph
Published
• View Publication
• BIB
Let $f_r(n)$ be the minimum number of complete $r$-partite $r$-graphs needed to partition the edge set of the complete $r$-uniform hypergraph on $n$ vertices. Graham and Pollak showed that $f_2(n) = n-1$. An easy construction shows that $f_r(n)\le (1-o(1))\binom{n}{\lfloor r/2\rfloor}$ and it has been unknown if this upper bound is asymptotically sharp. In this paper we show that $f_r(n)\le (\frac{14}{15}+o(1))\binom{n}{r/2}$ for each even $r\ge 4$.
Sets in Almost General Position
Erdős asked the following question: given $n$ points in the plane in almost general position (no 4 collinear), how large a set can we guarantee to find that is in general position (no 3 collinear)? Füredi constructed a set of $n$ points in almost general position with no more than $o(n)$ points in general position. Cardinal, Tóth and Wood extended this result to $\mathbb{R}^3$, finding sets of $n$ points with no 5 on a plane whose subsets with no 4 points on a plane have size $o(n)$, and asked the question for higher dimensions: for given $n$, is it still true that the largest subset in general position we can guarantee to find has size $o(n)$? We answer their question for all $d$ and derive improved bounds for certain dimensions.