Abstract:Invariance transformations of polyadic decompositions of matrix multiplication tensors define an equivalence relation on the set of such decompositions. In this paper, we present an algorithm to efficiently decide whether two polyadic decompositions of a given matrix multiplication tensor are equivalent. With this algorithm, we analyze the equivalence classes of decompositions of several matrix multiplication tensors. This analysis is relevant for the study of fast matrix multiplication as it relates to the question of how many essentially different fast matrix multiplication algorithms there exist. This question has been first studied by de~Groote, who showed that for the multiplication of $2\times2$ matrices with $7$ active multiplications, all algorithms are essentially equivalent to Strassen's algorithm. In contrast, the results of our analysis show that for the multiplication of larger matrices, (e.g., $2\times3$ by $3\times2$ or $3\times3$ by $3\times3$ matrices), two decompositions are very likely to be essentially different. We further provide a necessary criterion for a polyadic decomposition to be equivalent to a polyadic decomposition with integer entries. Decompositions with specific integer entries, e.g., powers of two, provide fast matrix multiplication algorithms with better efficiency and stability properties. This condition can be tested algorithmically and we present the conclusions obtained for the decompositions of small/medium matrix multiplication tensors.

Transference inequalities for multiplicative Diophantine exponents

On Diophantine exponents and Khintchine's transference principle

An analogue of Mahler's transference theorem for multiplicative Diophantine approximation

How to split Dyson's transference inequality with the help of Wolfgang Schmidt's parametric geometry of numbers

Independence of the Diophantine exponents associated with linear subspaces

Intermediate Diophantine exponents and parametric geometry of numbers

A simple proof of Schmidt-Summerer's inequality

Computing exponentials of essentially non-negative matrices entrywise to high relative accuracy.

Equivalent Polyadic Decompositions of Matrix Multiplication Tensors

Quantitative inhomogeneous Diophantine approximation for systems of linear forms

Exponential Erdős-Szekeres theorem for matrices

Ando-Hiai type inequalities for multivariate operator means

Integer matrices with a given characteristic polynomial and multiplicative dependence of matrices

Diophantine tuples and multiplicative structure of shifted multiplicative subgroups

On the extension for Toeplitz matrices of certain Markov inequalities

Counting and Joint equidistribution of approximates

Inequalities on $ 2\times 2 $ block accretive partial transpose matrices

Explicit inverse of symmetric, tridiagonal near Toeplitz matrices Part II: with weakly diagonally dominant Toeplitz

Dynamical diophantine approximation exponents in characteristic p

Quantitative lower bounds on the Lyapunov exponent from multivariate matrix inequalities

Exponential inequalities in probability spaces revisited