arXiv subjects
Publications and source records attributed to Alexandre Kosyak.
We show that $\frac{\Gamma(f_0,f_1,\dots,f_m)} {\Gamma(f_1,\dots,f_m)}=\infty$ for $m+1$ vectors having the properties that no non-trivial linear combination of them belongs to $l_2(\mathbb N)$. This property is essential in the proof of the irreducibility of unitary representations of some infinite-dimensional groups.
Consider the inductive limit of the general linear groups ${\rm GL}_0(2\infty,{\mathbb R})$ $= \varinjlim_{n}{\rm GL}(2n-1,{\mathbb R})$, acting on the space $X_m$ of $m$ rows, infinite in both directions, with Gaussian measure. This measure is the infinite tensor product of one-dimensional arbitrary Gaussian non-centered measures. In this article we prove an irreducibility criterion for $m=3$. In 2019, the first author [28] established a criterion for $m\le 2$. Our proof is in the same spirit, but the details are far more involved.
Given any positive integer $n,$ let $A(n)$ denote the height of the $n^{\text{th}}$ cyclotomic polynomial, that is its maximum coefficient in absolute value. It is well known that $A(n)$ is unbounded. We conjecture that every natural number can arise as value of $A(n)$ and prove this assuming that for every pair of consecutive primes $p$ and $p'$ with $p\ge 127$ we have $p'-p<\sqrt{p}+1.$ We also conjecture that every natural number occurs as maximum coefficient of some cyclotomic polynomial and show that this is true if Andrica's conjecture that always $\sqrt{p'}-\sqrt{p}<1$ holds. This is the first time, as far as the authors know, a connection between prime gaps and cyclotomic polynomials is uncovered. Using a result of Heath-Brown on prime gaps we show unconditionally that every natural number $m\le x$ occurs as $A(n)$ value with at most $O_ε(x^{3/5+ε})$ exceptions. On the Lindelöf Hypothesis we show there are at most $O_ε(x^{1/2+ε})$ exceptions and study them further by using deep work of Bombieri--Friedlander--Iwaniec on the distribution of primes in arithmetic progressions beyond the square-root barrier.
Our aim is to find the irreducibility criteria for the Koopman representation, when the group acts on some space with a measure (Conjecture 1.5). Some general necessary conditions of the irreducibility of this representation are established. In the particular case of the group ${\rm GL}_0(2\infty,{\mathbb R})$ $= \varinjlim_{n}{\rm GL}(2n-1,{\mathbb R})$, the inductive limit of the general linear groups we prove that these conditions are also the necessary ones. The corresponding measure is infinite tensor products of one-dimensional arbitrary Gaussian non-centered measures. The corresponding $G$-space $X_m$ is a subspace of the space ${\rm Mat}(2\infty,{\mathbb R})$ of infinite in both directions real matrices. In fact, $X_m$ is a collection of $m$ infinite in both directions rows. This result was announced in [20]. We give the proof only for $m\leq 2$. The general case will be studied later.
We construct the so-called quasiregular representations of the group $B_0^{\mathbb N}({\mathbb F}_p)$ of infinite upper triangular matrices with coefficients in a finite field and give the criteria of theirs irreducibility in terms of the initial measure. These representations are particular case of the Koopman representation hence, we find new conditions of its irreducibility. Since the field ${\mathbb F}_p$ is compact some new operators in the commutant emerges. Therefore, the Ismagilov conjecture in the case of the finite field should be corrected.
We show that the Lawrence--Krammer representation can be obtained as the quantization of the symmetric square of the Burau representation. This connection allows us to construct new representations of braid groups
The induced representation ${\rm Ind}_H^GS$ of a locally compact group $G$ is the unitary representation of the group $G$ associated with unitary representation $S:H\rightarrow U(V)$ of a subgroup $H$ of the group $G$. Our aim is to develop the concept of induced representations for infinite-dimensional groups. The induced representations for infinite-dimensional groups in not unique, as in the case of a locally compact groups. It depends on two completions $\tilde H$ and $\tilde G$ of the subgroup $H$ and the group $G$, on an extension $\tilde S:\tilde H\rightarrow U(V)$ of the representation $S:H\rightarrow U(V)$ and on a choice of the $G$-quasi-invariant measure $μ$ on an appropriate completion $\tilde X=\tilde H\backslash \tilde G$ of the space $H\backslash G$. As the illustration we consider the "nilpotent" group $B_0^{\mathbb Z}$ of infinite in both directions upper triangular matrices and the induced representation corresponding to the so-called generic
We study the von Neumann algebra, generated by the regular representations of the infinite-dimensional nilpotent group $B_0^{\mathbb Z}$. In [14] a condition have been found on the measure for the right von Neumann algebra to be the commutant of the left one. In the present article, we prove that, in this case, the von Neumann algebra generated by the regular representations of group $B_0^{\mathbb Z}$ is the type ${\rm III}_1$ hyperfinite factor. We use a technique, developed in [20] where a similar result was proved for the group $B_0^{\mathbb N}$. The crossed product allows us to remove some technical condition on the measure used in [20]. [14] A.V. Kosyak, Inversion-quasi-invariant Gaussian measures on the group of infinite-order upper-triangular matrices, Funct. Anal. i Priloz. 34, issue 1 (2000) 86--90. [20] A.V. Kosyak, Type ${\rm III_1}$ factors generated by regular representations of infinite dimensional nilpotent group $B_0^{\mathbb N}$, arXiv:0803.3340v1.
We give the criterion for the irreducibility, the Schur irreducibility and the indecomposability of the set of two $n\times n$ matrices $Λ_n$ and $A_n$ in terms of the subalgebra associated with the "support" of the matrix $A_n$, where $Λ_n$ is a diagonal matrix with different non zeros eigenvalues and $A_n$ is an arbitrary one. The list of all maximal subalgebras of the algebra ${\rm Mat}(n,{\mathbb C})$ and the list of the corresponding invariant subspaces connected with these two matrices is also given. The properties of the corresponding subalgebras are expressed in terms of the graphs associated with the support of the second matrix. For arbitrary $n$ we describe all minimal subsets of the elementary matrices $E_{km}$ that generate the algebra ${\rm Mat}(n,{\mathbb C})$.
We study the von Neumann algebra, generated by the unitary representations of infinite-dimensional groups nilpotent group $B_0^{\mathbb N}$. The conditions of the irreducibility of the regular and quasiregular representations of infinite-dimensional groups (associated with some quasi-invariant measures) are given by the so-called Ismagilov conjecture (see [1,2,9-11]). In this case the corresponding von Neumann algebra is type ${\rm I}_\infty$ factor. When the regular representation is reducible we find the sufficient conditions on the measure for the von Neumann algebra to be factor (see [13,14]). In the present article we determine the type of corresponding factors. Namely we prove that the von Neumann algebra generated by the regular representations of infinite-dimensional nilpotent group $B_0^{\mathbb N}$ is type ${\rm III}_1$ hyperfinite factor. The case of the nilpotent group $B_0^{\mathbb Z}$ of infinite in both directions matrices will be studied in [6].
We construct a [(n+1)/2]+1 parameters family of irreducible representations of the Braid group B_3 in arbitrary dimension n\in N, using a q-deformation of the Pascal triangle. This construction extends in particular results by S.P.Humphries [8], who constructed representations of the braid group B_3 in arbitrary dimension using the classical Pascal triangle. E.Ferrand [7] obtained an equivalent representation of B_3 by considering two special operators in the space C^n[X]. Slightly more general representations were given by I.Tuba and H.Wenzl [11]. They involve [(n+1)/2] parameters (and also use the classical Pascal triangle). The latter authors also gave the complete classification of all simple representations of B_3 for dimension n\leq 5. Our construction generalize all mentioned results and throws a new light on some of them. We also study the irreducibility and the equivalence of the representations. In [17] we establish the connection between the constructed representation of the braid group B_3 and the highest weight modules of U(sl_2) and quantum group U_q(sl_2).