It is shown that high order feedforward neural nets of constant depth with piecewise

polynomial activation functions and arbitrary real weights can be simulated for boolean

inputs and outputs by neural nets of a somewhat larger size and depth with heaviside

gates and weights ...
more >>>

Frederic Green

We say an integer polynomial $p$, on Boolean inputs, weakly

$m$-represents a Boolean function $f$ if $p$ is non-constant and is zero (mod

$m$), whenever $f$ is zero. In this paper we prove that if a polynomial

weakly $m$-represents the Mod$_q$ function on $n$ inputs, where $q$ and $m$

more >>>

Pascal Koiran

The main result of this paper is a Omega(n^{1/4}) lower bound

on the size of a sigmoidal circuit computing a specific AC^0_2 function.

This is the first lower bound for the computation model of sigmoidal

circuits with unbounded weights. We also give upper and lower bounds for

the ...
more >>>

Manindra Agrawal, Eric Allender

We show that all sets complete for NC$^1$ under AC$^0$

reductions are isomorphic under AC$^0$-computable isomorphisms.

Although our proof does not generalize directly to other

complexity classes, we do show that, for all complexity classes C

closed under NC$^1$-computable many-one reductions, the sets ...
more >>>

Shiva Chaudhuri, Jaikumar Radhakrishnan

We study the complexity of computing Boolean functions using AND, OR

and NOT gates. We show that a circuit of depth $d$ with $S$ gates can

be made to output a constant by setting $O(S^{1-\epsilon(d)})$ (where

$\epsilon(d) = 4^{-d}$) of its input values. This implies a

superlinear size lower bound ...
more >>>

Eric Allender

A very recent paper by Caussinus, McKenzie, Therien, and Vollmer

[CMTV] shows that ACC^0 is properly contained in ModPH, and TC^0

is properly contained in the counting hierarchy. Thus, [CMTV] shows

that there are problems in ModPH that require superpolynomial-size

uniform ACC^0 ...
more >>>

Eric Allender, Robert Beals, Mitsunori Ogihara

We characterize the complexity of some natural and important

problems in linear algebra. In particular, we identify natural

complexity classes for which the problems of (a) determining if a

system of linear equations is feasible and (b) computing the rank of

an integer matrix, ...
more >>>

Meera Sitharam

We develop an analytic framework based on

linear approximation and point out how a number of complexity

related questions --

on circuit and communication

complexity lower bounds, as well as

pseudorandomness, learnability, and general combinatorics

of Boolean functions --

fit neatly into this framework. ...
more >>>

Oded Goldreich, Avi Wigderson

We consider the size of circuits which perfectly hash

an arbitrary subset $S\!\subset\!\bitset^n$ of cardinality $2^k$

into $\bitset^m$.

We observe that, in general, the size of such circuits is

exponential in $2k-m$,

and provide a matching upper bound.

Per Enflo, Meera Sitharam

--

Scalar product estimates have so far been used in

proving several unweighted threshold lower bounds.

We show that if a basis set of Boolean functions satisfies

certain weak stability conditions, then

scalar product estimates

yield lower bounds for the size of weighted thresholds

of these basis functions.

Stable ...
more >>>

Alexander E. Andreev, Andrea E. F. Clementi, Jose' D. P. Rolim

We present the first worst-case hardness conditions

on the circuit complexity of EXP functions which are

sufficient to obtain P=BPP. In particular, we show that

from such hardness conditions it is possible to construct

quick Hitting Sets Generators with logarithmic prize.

...
more >>>

Anna Bernasconi, Igor E. Shparlinski

In this paper we extend the area of applications of

the Abstract Harmonic Analysis to the field of Boolean function complexity.

In particular, we extend the class of functions to which

a spectral technique developed in a series of works of the first author

can be applied.

This extension ...
more >>>

Paul Beame

Proof complexity, the study of the lengths of proofs in

propositional logic, is an area of study that is fundamentally connected

both to major open questions of computational complexity theory and

to practical properties of automated theorem provers. In the last

decade, there have been a number of significant advances ...
more >>>

Rüdiger Reischuk

For ordinary circuits with a fixed upper bound on the maximal fanin

of gates it has been shown that logarithmic redundancy is necessary and

sufficient to overcome random hardware faults.

Here, we consider the same question for unbounded fanin circuits that

in the noiseless case can compute ...
more >>>

Ronen Shaltiel

A fundamental question of complexity theory is the direct product

question. Namely weather the assumption that a function $f$ is hard on

average for some computational class (meaning that every algorithm from

the class has small advantage over random guessing when computing $f$)

entails that computing $f$ on ...
more >>>

Hubie Chen

Representations of boolean functions as polynomials (over rings) have

been used to establish lower bounds in complexity theory. Such

representations were used to great effect by Smolensky, who

established that MOD q \notin AC^0[MOD p] (for distinct primes p, q)

by representing AC^0[MOD p] functions as low-degree multilinear

polynomials over ...
more >>>

Robert Albin Legenstein

We introduce em total wire length as salient complexity measure

for analyzing the circuit complexity of sensory processing in

biological neural systems and neuromorphic engineering. The new

complexity measure is applied in this paper to two basic

computational problems that arise in translation- and

scale-invariant pattern recognition, and hence appear ...
more >>>

Robert Albin Legenstein

One of the most basic pattern recognition problems is whether a

certain local feature occurs in some linear array to the left of

some other local feature. We construct in this article circuits that

solve this problem with an asymptotically optimal number of

threshold gates. Furthermore it is shown that ...
more >>>

Hubie Chen

The boolean circuit complexity classes

AC^0 \subseteq AC^0[m] \subseteq TC^0 \subseteq NC^1 have been studied

intensely. Other than NC^1, they are defined by constant-depth

circuits of polynomial size and unbounded fan-in over some set of

allowed gates. One reason for interest in these classes is that they

contain the ...
more >>>

Ran Raz

An arithmetic formula is multi-linear if the polynomial computed

by each of its sub-formulas is multi-linear. We prove that any

multi-linear arithmetic formula for the permanent or the

determinant of an $n \times n$ matrix is of size super-polynomial

in $n$.

Ran Raz

An arithmetic circuit or formula is multilinear if the polynomial

computed at each of its wires is multilinear.

We give an explicit example for a polynomial $f(x_1,...,x_n)$,

with coefficients in $\{0,1\}$, such that over any field:

1) $f$ can be computed by a polynomial-size multilinear circuit

of depth $O(\log^2 ...
more >>>

Eric Allender, Harry Buhrman, Michal Koucky

We investigate the question of whether one can characterize complexity

classes (such as PSPACE or NEXP) in terms of efficient

reducibility to the set of Kolmogorov-random strings R_C.

We show that this question cannot be posed without explicitly dealing

with issues raised by the choice of universal

machine in the ...
more >>>

Xiaoyang Gu

In this paper, we use resource-bounded dimension theory to investigate polynomial size circuits. We show that for every $i\geq 0$, $\Ppoly$ has $i$th order scaled $\pthree$-strong dimension $0$. We also show that $\Ppoly^\io$ has $\pthree$-dimension $1/2$, $\pthree$-strong dimension $1$. Our results improve previous measure results of Lutz (1992) and dimension ... more >>>

Vinodchandran Variyam

In this short note we show that for any integer k, there are

languages in the complexity class PP that do not have Boolean

circuits of size $n^k$.

Eric Allender, Samir Datta, Sambuddha Roy

We show that ACC^0 is precisely what can be computed with constant-width circuits of polynomial size and polylogarithmic genus. This extends a characterization given by Hansen, showing that planar constant-width circuits also characterize ACC^0. Thus polylogarithmic genus provides no additional computational power in this model.

We consider other generalizations of ...
more >>>

Oded Goldreich

The notion of promise problems was introduced and initially studied

by Even, Selman and Yacobi

(Information and Control, Vol.~61, pages 159-173, 1984).

In this article we survey some of the applications that this

notion has found in the twenty years that elapsed.

These include the notion ...
more >>>

Joan Boyar, rene peralta

We consider the problem of computing the Hamming weight of an n-bit vector using a circuit with gates for GF2 addition and multiplication only. We show the number of multiplications necessary and sufficient to build such a circuit is n - |n| where |n| is the Hamming weight of the ... more >>>

Mahdi Cheraghchi

The rigidity function of a matrix is defined as the minimum number of its entries that need to be changed in order to reduce the rank of the matrix to below a given parameter. Proving a strong enough lower bound on the rigidity of a matrix implies a nontrivial lower ... more >>>

Guy Wolfovitz

We give tight lower bounds for the size of depth-3 circuits with limited bottom fanin computing symmetric Boolean functions. We show that any depth-3 circuit with bottom fanin $k$ which computes the Boolean function $EXACT_{n/(k+1)}^{n}$, has at least $(1+1/k)^{n+\O(\log n)}$ gates. We show that this lower bound is tight, by ... more >>>

Arkadev Chattopadhyay

Let m,q > 1 be two integers that are co-prime and A be any subset of Z_m. Let P be any multi-linear polynomial of degree d in n variables over Z_m. We show that the MOD_q boolean function on n variables has correlation at most exp(-\Omega(n/(m2^{m-1})^d)) with the boolean function ... more >>>

Kei Uchizawa, Rodney Douglas

Circuits composed of threshold gates (McCulloch-Pitts neurons, or

perceptrons) are simplified models of neural circuits with the

advantage that they are theoretically more tractable than their

biological counterparts. However, when such threshold circuits are

designed to perform a specific computational task they usually

differ ...
more >>>

Eric Allender, Michal Koucky

We observe that many important computational problems in NC^1 share a simple self-reducibility property. We then show that, for any problem A having this self-reducibility property, A has polynomial size TC^0 circuits if and only if it has TC^0 circuits of size n^{1+\epsilon} for every \epsilon > 0 (counting the ... more >>>

Mark Braverman

We prove that poly-sized AC0 circuits cannot distinguish a poly-logarithmically independent distribution from the uniform one. This settles the 1990 conjecture by Linial and Nisan [LN90]. The only prior progress on the problem was by Bazzi [Baz07], who showed that O(log^2 n)-independent distributions fool poly-size DNF formulas. Razborov [Raz08] has ... more >>>

Eric Allender, Michal Koucky, Detlef Ronneburger, Sambuddha Roy

We continue an investigation into resource-bounded Kolmogorov complexity \cite{abkmr}, which highlights the close connections between circuit complexity and Levin's time-bounded Kolmogorov complexity measure Kt (and other measures with a similar flavor), and also exploits derandomization techniques to provide new insights regarding Kolmogorov complexity.

The Kolmogorov measures that have been ...
more >>>

Christoph Behle, Andreas Krebs, Stephanie Reifferscheid

We consider the regular languages recognized by weighted threshold circuits with a linear number of wires.

We present a simple proof to show that parity cannot be computed by such circuits.

Our proofs are based on an explicit construction to restrict the input of the circuit such that the value ...
more >>>

Dan Gutfreund, Akinori Kawachi

We show that if Arthur-Merlin protocols can be derandomized, then there is a Boolean function computable in deterministic exponential-time with access to an NP oracle, that cannot be computed by Boolean circuits of exponential size. More formally, if $\mathrm{prAM}\subseteq \mathrm{P}^{\mathrm{NP}}$ then there is a Boolean function in $\mathrm{E}^{\mathrm{NP}}$ that requires ... more >>>

Christoph Behle, Andreas Krebs, Klaus-Joern Lange, Pierre McKenzie

In the setting known as DLOGTIME-uniformity,

the fundamental complexity classes

$AC^0\subset ACC^0\subseteq TC^0\subseteq NC^1$ have several

robust characterizations.

In this paper we refine uniformity further and examine the impact

of these refinements on $NC^1$ and its subclasses.

When applied to the logarithmic circuit depth characterization of $NC^1$,

some refinements leave ...
more >>>

Michael Elberfeld, Andreas Jakoby, Till Tantau

An algorithmic meta theorem for a logic and a class $C$ of structures states that all problems expressible in this logic can be solved efficiently for inputs from $C$. The prime example is Courcelle's Theorem, which states that monadic second-order (MSO) definable problems are linear-time solvable on graphs of bounded ... more >>>

Alexander A. Sherstov

We study the set disjointness problem in the number-on-the-forehead model.

(i) We prove that $k$-party set disjointness has randomized and nondeterministic

communication complexity $\Omega(n/4^k)^{1/4}$ and Merlin-Arthur complexity $\Omega(n/4^k)^{1/8}.$

These bounds are close to tight. Previous lower bounds (2007-2008) for $k\geq3$ parties

were weaker than $n^{1/(k+1)}/2^{k^2}$ in all ...
more >>>

Matthew Anderson, Dieter van Melkebeek, Nicole Schweikardt, Luc Segoufin

We study the locality of an extension of first-order logic that captures graph queries computable in AC$^0$, i.e., by families of polynomial-size constant-depth circuits. The extension considers first-order formulas over relational structures which may use arbitrary numerical predicates in such a way that their truth value is independent of the ... more >>>

Rahul Santhanam, Ryan Williams

We explore the relationships between circuit complexity, the complexity of generating circuits, and circuit-analysis algorithms. Our results can be roughly divided into three parts:

1. Lower Bounds Against Medium-Uniform Circuits. Informally, a circuit class is ``medium uniform'' if it can be generated by an algorithmic process that is somewhat complex ... more >>>

Kristoffer Arnsfelt Hansen, Vladimir Podolskii

We study the complexity of computing Boolean functions on general

Boolean domains by polynomial threshold functions (PTFs). A typical

example of a general Boolean domain is $\{1,2\}^n$. We are mainly

interested in the length (the number of monomials) of PTFs, with

their degree and weight being of secondary interest. We ...
more >>>

Andreas Krebs, Nutan Limaye, Meena Mahajan, Karteek Sreenivasaiah

A proof system for a language $L$ is a function $f$ such that Range$(f)$ is exactly $L$. In this paper, we look at proofsystems from a circuit complexity point of view and study proof systems that are computationally very restricted. The restriction we study is: they can be computed by ... more >>>

Olaf Beyersdorff, Leroy Chew, Mikolas Janota

Proof systems for quantified Boolean formulas (QBFs) provide a theoretical underpinning for the performance of important

QBF solvers. However, the proof complexity of these proof systems is currently not well understood and in particular

lower bound techniques are missing. In this paper we exhibit a new and elegant proof technique ...
more >>>

Eric Allender, Ian Mertz

We give complexity bounds for various classes of functions computed by cost register automata.

more >>>Mahdi Cheraghchi, Elena Grigorescu, Brendan Juba, Karl Wimmer, Ning Xie

$\mathrm{AC}^{0} \circ \mathrm{MOD}_2$ circuits are $\mathrm{AC}^{0}$ circuits augmented with a layer of parity gates just above the input layer. We study the $\mathrm{AC}^{0} \circ \mathrm{MOD}_2$ circuit lower bound for computing the Boolean Inner Product functions. Recent works by Servedio and Viola (ECCC TR12-144) and Akavia et al. (ITCS 2014) have ... more >>>

Daniel Kane, Ryan Williams

In order to formally understand the power of neural computing, we first need to crack the frontier of threshold circuits with two and three layers, a regime that has been surprisingly intractable to analyze. We prove the first super-linear gate lower bounds and the first super-quadratic wire lower bounds for ... more >>>

Alexander Golovnev, Alexander Kulikov, Alexander Smal, Suguru Tamaki

Most of the known lower bounds for binary Boolean circuits with unrestricted depth are proved by the gate elimination method. The most efficient known algorithms for the #SAT problem on binary Boolean circuits use similar case analyses to the ones in gate elimination. Chen and Kabanets recently showed that the ... more >>>

Jan Krajicek, Igor Carboni Oliveira

We establish unconditionally that for every integer $k \geq 1$ there is a language $L \in P$ such that it is consistent with Cook's theory PV that $L \notin SIZE(n^k)$. Our argument is non-constructive and does not provide an explicit description of this language.

more >>>Arkadev Chattopadhyay, Nikhil Mande

We show that a simple function has small unbounded error communication complexity in the $k$-party number-on-forehead (NOF) model but every probabilistic protocol that solves it with sub-exponential advantage over random guessing has cost essentially $\Omega\left(\frac{\sqrt{n}}{4^k}\right)$ bits. Such a separation was first shown for $k=2$ independently by Buhrman et al. ['07] ... more >>>

Alexander Golovnev, Oded Regev, Omri Weinstein

The minrank of a graph $G$ is the minimum rank of a matrix $M$ that can be obtained from the adjacency matrix of $G$ by switching ones to zeros (i.e., deleting edges) and setting all diagonal entries to one. This quantity is closely related to the fundamental information-theoretic problems of ... more >>>

Alexander Kulikov, Vladimir Podolskii

We study the following computational problem: for which values of $k$, the majority of $n$ bits $\text{MAJ}_n$ can be computed with a depth two formula whose each gate computes a majority function of at most $k$ bits? The corresponding computational model is denoted by $\text{MAJ}_k \circ \text{MAJ}_k$. We observe that ... more >>>

Andreas Krebs, Nutan Limaye, Michael Ludwig

In this work we consider the term evaluation problem which involves, given a term over some algebra and a valid input to the term, computing the value of the term on that input. This is a classical problem studied under many names such as formula evaluation problem, formula value problem ... more >>>

Benjamin Rossman, Srikanth Srinivasan

This paper gives the first separation between the power of {\em formulas} and {\em circuits} of equal depth in the $\mathrm{AC}^0[\oplus]$ basis (unbounded fan-in AND, OR, NOT and MOD$_2$ gates). We show, for all $d(n) \le O(\frac{\log n}{\log\log n})$, that there exist {\em polynomial-size depth-$d$ circuits} that are not equivalent ... more >>>

Shuichi Hirahara

We establish an explicit link between depth-3 formulas and one-sided approximation by depth-2 formulas, which were previously studied independently. Specifically, we show that the minimum size of depth-3 formulas is (up to a factor of n) equal to the inverse of the maximum, over all depth-2 formulas, of one-sided-error correlation ... more >>>

Russell Impagliazzo, Valentine Kabanets, Antonina Kolokolova, Pierre McKenzie, Shadab Romani

The Black-Box Hypothesis, introduced by Barak et al. (JACM, 2012), states that any property of boolean functions decided efficiently (e.g., in BPP) with inputs represented by circuits can also be decided efficiently in the black-box setting, where an algorithm is given an oracle access to the input function and an ... more >>>

Or Meir, Avi Wigderson

Consider a random sequence of $n$ bits that has entropy at least $n-k$, where $k\ll n$. A commonly used observation is that an average coordinate of this random sequence is close to being uniformly distributed, that is, the coordinate “looks random”. In this work, we prove a stronger result that ... more >>>

Cody Murray, Ryan Williams

We prove that if every problem in $NP$ has $n^k$-size circuits for a fixed constant $k$, then for every $NP$-verifier and every yes-instance $x$ of length $n$ for that verifier, the verifier's search space has an $n^{O(k^3)}$-size witness circuit: a witness for $x$ that can be encoded with a circuit ... more >>>

Alexander Smal, Navid Talebanfard

Let $X$ be a random variable distributed over $n$-bit strings with $H(X) \ge n - k$, where $k \ll n$. Using subadditivity we know that a random coordinate looks random. Meir and Wigderson [TR17-149] showed a random coordinate looks random to an adversary who is allowed to query around $n/k$ ... more >>>

Aayush Ojha, Raghunath Tewari

Recently, perfect matching in bounded planar cutwidth bipartite graphs

$BGGM$ was shown to be in ACC$^0$ by Hansen et al.. They also conjectured that

the problem is in AC$^0$.

In this paper, we disprove their conjecture by showing that the problem is

not in AC$^0[p^{\alpha}]$ for every prime $p$. ...
more >>>

Igor Carboni Oliveira, Rahul Santhanam

We show that for several natural problems of interest, complexity lower bounds that are barely non-trivial imply super-polynomial or even exponential lower bounds in strong computational models. We term this phenomenon "hardness magnification". Our examples of hardness magnification include:

1. Let MCSP$[s]$ be the decision problem whose YES instances are ... more >>>

Igor Carboni Oliveira

We introduce randomized time-bounded Kolmogorov complexity (rKt), a natural extension of Levin's notion of Kolmogorov complexity from 1984. A string w of low rKt complexity can be decompressed from a short representation via a time-bounded algorithm that outputs w with high probability.

This complexity measure gives rise to a ... more >>>

Igor Carboni Oliveira, Rahul Santhanam, Srikanth Srinivasan

We study the complexity of computing symmetric and threshold functions by constant-depth circuits with Parity gates, also known as AC$^0[\oplus]$ circuits. Razborov (1987) and Smolensky (1987, 1993) showed that Majority requires depth-$d$ AC$^0[\oplus]$ circuits of size $2^{\Omega(n^{1/2(d-1)})}$. By using a divide-and-conquer approach, it is easy to show that Majority can ... more >>>

Jan Bydzovsky, Igor Carboni Oliveira, Jan Krajicek

Proving that there are problems in $P^{NP}$ that require boolean circuits of super-linear size is a major frontier in complexity theory. While such lower bounds are known for larger complexity classes, existing results only show that the corresponding problems are hard on infinitely many input lengths. For instance, proving almost-everywhere ... more >>>

Anna Gal, Robert Robere

Comparator circuits are a natural circuit model for studying the concept of bounded fan-out computations, which intuitively corresponds to whether or not a computational model can make "copies" of intermediate computational steps. Comparator circuits are believed to be weaker than general Boolean circuits, but they can simulate Branching Programs and ... more >>>

Yaroslav Alekseev, Dima Grigoriev, Edward Hirsch, Iddo Tzameret

We introduce the `binary value principle' which is a simple subset-sum instance expressing that a natural number written in binary cannot be negative, relating it to central problems in proof and algebraic complexity. We prove conditional superpolynomial lower bounds on the Ideal Proof System (IPS) refutation size of this instance, ... more >>>

Igor Carboni Oliveira, Lijie Chen, Shuichi Hirahara, Ján Pich, Ninad Rajgopal, Rahul Santhanam

Hardness magnification reduces major complexity separations (such as $EXP \not\subseteq NC^1$) to proving lower bounds for some natural problem $Q$ against weak circuit models. Several recent works [OS18, MMW19, CT19, OPS19, CMMW19, Oli19, CJW19a] have established results of this form. In the most intriguing cases, the required lower bound is ... more >>>