arXiv Science⌕ Search
Home/Archive/arXiv

arXiv · 2609.28839

Exponential Correlation Bounds for Polynomials

Eshan Chattopadhyay·Pooya Hatami·Chin Ho Lee·Shachar Lovett·Avishay Tal·Emanuele Viola
View original source ↗Read source document ↗Export metadata ↓

Abstract

We prove that the XOR of $k$ majorities on disjoint blocks of $\ell$ bits has correlation at most $(4d/\sqrt{\ell})^k$ with every Boolean-valued degree-$d$ polynomial over an arbitrary field. By known techniques, we obtain explicit pseudorandom generators that fool such polynomials on $n$ Boolean inputs with seed length $d^2\cdot\polylog(n/\varepsilon)$. The correlation and the generator extend to small $\mathsf{AC}^0[q]$ circuits, for every fixed prime power $q$. These results resolve several long-standing questions in circuit complexity and pseudorandomness.

Explore related subjects

cs.CC

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Eshan Chattopadhyay, Pooya Hatami, Chin Ho Lee, Shachar Lovett, Avishay Tal, Emanuele Viola. 2026-10-11. Exponential Correlation Bounds for Polynomials. https://arxiv.org/abs/2609.28839

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

The Polynomial Hierarchy does not collapse

The arithmetical hierarchy (AH) is similar to the polynomial hierarchy (PH). Unlike the PH, the AH does not collapse relative to any oracle. A $Σ_{k+1}$-complete language is computably enumerable (c.e.) relative to the k-th level of the AH. So, given an oracle in the k-th level of the AH, we require exhaustive search to recognize whether the input string is in the language. We can define a language in AH with a quantified formula and a computable relation. The encoding lengths of the bounded variables are unbounded. However, if we observe only finitely many input strings, then there exist integers such that the encoding lengths of the bounded variables can be limited to these integers without changing the result of the formula for each input string. By adding padding arguments of length equal to those of the bounded variables and padding of length equal to the runtime required to compute the relation, we construct an equivalent language in PH. An OTM operating on the language with padding arguments in PH with a padded oracle has the same program flow as it does on the original language with the unpadded oracle in the AH. The number of oracle queries does not change. We still need brute-force search therefore, the required number of oracle queries is exponential. The PH does not collapse.

cs.CC↗
arXiv2026

PSPACE-completeness of the Hunters and Rabbit game

In the Hunters and Rabbit game, hunters shoot at a set of vertices at each round, while an invisible rabbit must move along an edge. We prove that deciding whether a given number of hunters can guarantee capture is PSPACE-complete, even on simple bipartite graphs. The reduction uses intersection nonemptiness for inverse finite automata: one common word must put all automata in accepting states simultaneously. A clique acting as a safe haven and two families of guards force every winning strategy to simulate common transitions. Existing transformations remove the restriction on the rabbit's initial positions and yield PSPACE-completeness for the directed game with one hunter, even on oriented graphs with positive minimum outdegree. Adjusting the guard sizes also proves PSPACE-completeness of the firefighting game, in which the rabbit may remain at its current vertex.

cs.CC↗
arXiv2026

The Separation of $\mathit{NP}$ and $\mathit{PSPACE}$

There is an important and interesting open question in computational complexity on the relation between the complexity classes $\mathcal{NP}$ and $\mathcal{PSPACE}$. It is a widespread belief that $\mathcal{NP}\ne\mathcal{PSPACE}$. In this paper, we confirm this conjecture affirmatively by showing that there is a language $L_d$ accepted by no polynomial-time nondeterministic Turing machines but accepted by a nondeterministic Turing machine running within space $O(n^k)$ for all $k\in\mathbb{N}_1$. We achieve this by virtue of the prerequisite of $$ {\rm NTIME}[S(n)]\subseteq{\rm DSPACE}[S(n)], $$ and then by diagonalization against all polynomial-time nondeterministic Turing machines via a universal nondeterministic Turing machine $M_0$. We further show that $L_d\in \mathcal{PSPACE}$, which leads to the conclusion $$ \mathcal{NP}\subsetneqq\mathcal{PSPACE}. $$ Our approach is based on standard diagonalization and novel new techniques developed in the author's recent works \cite{Lin21a,Lin21b} with some new refinement.

cs.CC↗