AFTD Auto-Formalizing Theoretical Domains

matrix3_tensor_rank_z_le_23

theoremverifiedformalization

The tensor rank of 3x3 matrix multiplication over the integers is at most 23; that is, there exists a valid decomposition of 3x3 matrix multiplication over ℤ of length at most 23.

Statement

theorem matrix3_tensor_rank_z_le_23 : matrix3_tensor_rank_le ℤ 23
Formalizes
arXiv:2610.01639 Sec. 2.1 (citing A noncommutative algorithm for multiplying 3x3 matrices using 23 multiplications, 1976)
Verified
05 Oct 2026
Axioms
Classical.choiceQuot.soundpropext
Built on 2
  • matrix3_tensor_rank_leFor a commutative semiring R and a natural number r, matrix3_tensor_rank_le R r asserts that there exists a list L of elements of…
  • matrix3_tensor_rank_le_23_of_comm_ringOver every commutative ring R, 3x3 matrix multiplication has a valid bilinear decomposition with 23 products (the classical…

Read back from the Lean

There exists a list L of triads of 3×3 integer matrices (Matrix3Triad ℤ, consisting of matrices u, v, and w) such that L.length ≤ 23 and, for all 3×3 integer matrices A and B, the sum of eval_matrix3_triad t A B over all triads t in L equals the matrix product A * B.

Written by a model that saw only the Lean, never the English above. If the two disagree, that is worth a challenge.

Lean source view module on GitHub

/-- The upper bound of 23 multiplications for 3x3 matrix multiplication over the integers ℤ. -/
theorem matrix3_tensor_rank_z_le_23 : matrix3_tensor_rank_le ℤ 23 := matrix3_tensor_rank_le_23_of_comm_ring ℤ
See it in the graphDiscuss this resultChallenge it