A universal sequence of tensors for the asymptotic rank conjecture
Abstract
The exponent of a tensor over a field captures the base of the exponential growth rate of the tensor rank of under Kronecker powers. Tensor exponents are fundamental from the standpoint of algorithms and computational complexity theory; for example, the exponent of matrix multiplication can be characterized as , where is the tensor that represents matrix multiplication. Our main result is an explicit construction of a sequence of zero-one-valued tensors that is universal for the worst-case tensor exponent; more precisely, we show that where . We also supply an explicit universal sequence localised to capture the worst-case exponent of tensors with support contained in ; by combining such sequences, we obtain a universal sequence such that holds if and only if Strassen's asymptotic rank conjecture [Progr. Math. 120 (1994)] holds for . Finally, we show that the limit exists and can be captured as for an explicit sequence of tensors obtained by diagonalisation of the sequences . As our second result we relate the absence of polynomials of fixed degree vanishing on tensors of low rank, or more generally asymptotic rank, with upper bounds on the exponent . Using this technique, one may bound asymptotic rank for all tensors of a given format, knowing enough specific tensors of low asymptotic rank.
Keywords
Cite
@article{arxiv.2404.06427,
title = {A universal sequence of tensors for the asymptotic rank conjecture},
author = {Petteri Kaski and Mateusz Michałek},
journal= {arXiv preprint arXiv:2404.06427},
year = {2024}
}