New lower bounds for matrix multiplication and the 3x3 determinant

Abstract

Let $M_{\langle u,v,w\rangle}\in C^{uv}\otimes C^{vw}\otimes C^{wu}$ denote the matrix multiplication tensor (and write $M_n=M_{\langle n,n,n\rangle}$) and let $det_3\in ( C^9)^{\otimes 3}$ denote the determinant polynomial considered as a tensor. For a tensor $T$, let $\underline R(T)$ denote its border rank. We (i) give the first hand-checkable algebraic proof that $\underline R(M_2)=7$,(ii) prove $\underline R(M_{\langle 223\rangle})=10$, and $\underline R(M_{\langle 233\rangle})=14$, where previously the only nontrivial matrix multiplication tensor whose border rank had been determined was $M_2$,(iii) prove $\underline R( M_3)\geq 17$, (iv) prove $\underline R( det_3)=17$, improving the previous lower bound of $12$, (v) prove $\underline R(M_{\langle 2nn\rangle})\geq n^2+1.32n$ for all $n\geq 25$ (previously only $\underline R(M_{\langle 2nn\rangle})\geq n^2+1$ was known) as well as lower bounds for $4\leq n\leq 25$, and (vi) prove $\underline R(M_{\langle 3nn\rangle})\geq n^2+2 n+1$ for all $ n\geq 21$, where previously only $\underline R(M_{\langle 3nn\rangle})\geq n^2+2$ was known, as well as lower boundsfor $4\leq n\leq 21$. Our results utilize a new technique initiated by Buczyńska and Buczyński, called border apolarity. The two key ingredients are: (i) the use of a multi-graded ideal associated to a border rank $r$ decomposition of any tensor, and (ii) the exploitation of the large symmetry group of $T$ to restrict to $B_T$-invariant ideals, where $B_T$ is a maximal solvable subgroup of the symmetry group of $T$.

Abstract: arXiv descriptive metadata (CC0). Thank you to arXiv for use of its open access interoperability. This service was not reviewed or approved by, nor does it necessarily express or reflect the policies or opinions of, arXiv.