Prove2Me
Navigate
MissionsFormalpediaBlogsUsersMy Missions+
Prove2Me
⌕
Log in
← Formalpedia

Section 5 and Table 1: fourth-power support and ten classes

Proved
mme_stothers_fourth_support_and_classes

by marwahaha · Aug 29, 2026 · Mathlib 777aaa6 (Lean v4.29.0-rc3)

algebraic-complexitycoppersmith-winogradlaser-methodmatrix-multiplication

Let KKK be an arbitrary field. In the canonical nine-grading of the literal fourth power CW6⊗4CW_6^{\otimes4}CW6⊗4​:

  1. every block whose three mode grades do not sum to 888 has zero block tensor;
  2. exactly 454545 grade triples sum to 888;
  3. every such triple belongs to a unique permutation orbit represented by one of 008,017,026,035,044,116,125,134,224,233008,017,026,035,044,116,125,134,224,233008,017,026,035,044,116,125,134,224,233; and
  4. the orbit represented by class iii has cardinality 3ni3n_i3ni​, where n=(1,2,2,2,1,1,2,2,1,1)n=(1,2,2,2,1,1,2,2,1,1)n=(1,2,2,2,1,1,2,2,1,1).

This is the literal fourth-power decomposition underlying Section 5 and Table 1; it does not replace overlapping blocks by an external direct sum.

Preamble
import Definitions.Def_mme_stothers_fourth_data

open MME BigOperators

universe u

set_option autoImplicit false
Formal statement
theorem mme_stothers_fourth_support_and_classes
    {K : Type u} [Field K] :
    (∀ sigma : Fin 3 → Fin 9,
      (MME.StothersFourth.cwFourthCanonicalGrading K 6).blockTensor sigma = 0 ↔
        (∑ s, (sigma s).val) ≠ 8) ∧
    Fintype.card
        {sigma : Fin 3 → Fin 9 // (∑ s, (sigma s).val) = 8} = 45 ∧
    (∀ sigma : {sigma : Fin 3 → Fin 9 // (∑ s, (sigma s).val) = 8},
      ∃! r : Fin 10,
        MME.StothersFourth.sameOrbit sigma.1
          (MME.StothersFourth.classRep r)) ∧
    (∀ r : Fin 10,
      Fintype.card
          {sigma : {sigma : Fin 3 → Fin 9 // (∑ s, (sigma s).val) = 8} //
            MME.StothersFourth.sameOrbit sigma.1
              (MME.StothersFourth.classRep r)} =
        3 * MME.StothersFourth.classMultiplicity r) := by
  sorry
Source
Davie and Stothers (2013), Section 5, paragraph preceding Lemma 5.1 and Table 1, printed pp. 363 and 367, https://www.maths.ed.ac.uk/~sandy/a11164.pdf.
Read-back

What the Lean code literally says, in plain math · gpt-5

For an arbitrary universe uuu and every type KKK in that universe equipped with a field structure, with no further restriction on KKK, let Fin⁡(n)={0,…,n−1}\operatorname{Fin}(n)=\{0,\ldots,n-1\}Fin(n)={0,…,n−1}, put C=CWObj⁡(K,6)C=\operatorname{CWObj}(K,6)C=CWObj(K,6), and form the three-slot tensor object T=kron⁡(kron⁡(C,C),kron⁡(C,C))T=\operatorname{kron}(\operatorname{kron}(C,C),\operatorname{kron}(C,C))T=kron(kron(C,C),kron(C,C)). At each slot s∈Fin⁡(3)s\in\operatorname{Fin}(3)s∈Fin(3), the “fourth canonical” basis of TTT is, literally, the tensor-product basis Bs=Bs□⊗Bs□B_s=B_s^\square\otimes B_s^\squareBs​=Bs□​⊗Bs□​ of two copies of the square canonical basis Bs□B_s^\squareBs□​ for KKK and q=6q=6q=6; its indices are pairs p=(p1,p2)p=(p_1,p_2)p=(p1​,p2​) with p1,p2∈Fin⁡(8)×Fin⁡(8)p_1,p_2\in\operatorname{Fin}(8)\times\operatorname{Fin}(8)p1​,p2​∈Fin(8)×Fin(8). If g□g_\squareg□​ denotes the supplied square-pair grade at q=6q=6q=6, the fourth pair grade g4(p)∈Fin⁡(9)g_4(p)\in\operatorname{Fin}(9)g4​(p)∈Fin(9) is the element whose underlying natural number is val⁡(g□(p1))+val⁡(g□(p2))\operatorname{val}(g_\square(p_1))+\operatorname{val}(g_\square(p_2))val(g□​(p1​))+val(g□​(p2​)). The grading Γ\GammaΓ used below is the internal nine-part basis grading whose decomposition at each slot sss is the basis grading induced by BsB_sBs​ and g4g_4g4​; for a function σ:Fin⁡(3)→Fin⁡(9)\sigma:\operatorname{Fin}(3)\to\operatorname{Fin}(9)σ:Fin(3)→Fin(9), write TΓ[σ]T_\Gamma[\sigma]TΓ​[σ] for the block tensor selected by the three grades σ(0),σ(1),σ(2)\sigma(0),\sigma(1),\sigma(2)σ(0),σ(1),σ(2). Also define ten representative functions ρr:Fin⁡(3)→Fin⁡(9)\rho_r:\operatorname{Fin}(3)\to\operatorname{Fin}(9)ρr​:Fin(3)→Fin(9), displayed as ordered triples of their values at 0,1,20,1,20,1,2, and ten natural-number multiplicities mrm_rmr​ by (ρ0,…,ρ9)=((0,0,8),(0,1,7),(0,2,6),(0,3,5),(0,4,4),(1,1,6),(1,2,5),(1,3,4),(2,2,4),(2,3,3))(\rho_0,\ldots,\rho_9)=((0,0,8),(0,1,7),(0,2,6),(0,3,5),(0,4,4),(1,1,6),(1,2,5),(1,3,4),(2,2,4),(2,3,3))(ρ0​,…,ρ9​)=((0,0,8),(0,1,7),(0,2,6),(0,3,5),(0,4,4),(1,1,6),(1,2,5),(1,3,4),(2,2,4),(2,3,3)) and (m0,…,m9)=(1,2,2,2,1,1,2,2,1,1)(m_0,\ldots,m_9)=(1,2,2,2,1,1,2,2,1,1)(m0​,…,m9​)=(1,2,2,2,1,1,2,2,1,1). The theorem asserts all four of the following simultaneously: first, for every one of the 939^393 functions σ:Fin⁡(3)→Fin⁡(9)\sigma:\operatorname{Fin}(3)\to\operatorname{Fin}(9)σ:Fin(3)→Fin(9), including functions with repeated values and values 000 or 888, TΓ[σ]=0T_\Gamma[\sigma]=0TΓ​[σ]=0 if and only if the natural-number sum ∑s∈Fin⁡(3)val⁡(σ(s))\sum_{s\in\operatorname{Fin}(3)}\operatorname{val}(\sigma(s))∑s∈Fin(3)​val(σ(s)) is not 888; both directions are asserted—zero implies that the sum is not 888, and a sum different from 888 implies zero—so a sum equal to 888 forces the selected block tensor to be nonzero. Second, the set S={σ:Fin⁡(3)→Fin⁡(9)∣∑sval⁡(σ(s))=8}S=\{\sigma:\operatorname{Fin}(3)\to\operatorname{Fin}(9)\mid\sum_s\operatorname{val}(\sigma(s))=8\}S={σ:Fin(3)→Fin(9)∣∑s​val(σ(s))=8} has exactly 454545 elements. Third, for every σ∈S\sigma\in Sσ∈S, there exists exactly one class index r∈Fin⁡(10)r\in\operatorname{Fin}(10)r∈Fin(10) such that there exists a permutation eee of Fin⁡(3)\operatorname{Fin}(3)Fin(3) satisfying σ(s)=ρr(e(s))\sigma(s)=\rho_r(e(s))σ(s)=ρr​(e(s)) for every s∈Fin⁡(3)s\in\operatorname{Fin}(3)s∈Fin(3); this is the literal direction of the equality, and only rrr is unique—the permutation eee is merely existential and need not be unique, in particular when a representative has repeated coordinates. Fourth, for every r∈Fin⁡(10)r\in\operatorname{Fin}(10)r∈Fin(10), the number of distinct σ∈S\sigma\in Sσ∈S for which such a permutation eee exists is exactly 3mr3m_r3mr​, hence is respectively (3,6,6,6,3,3,6,6,3,3)(3,6,6,6,3,3,6,6,3,3)(3,6,6,6,3,3,6,6,3,3) for r=0,…,9r=0,\ldots,9r=0,…,9; this cardinality counts the functions σ\sigmaσ, not the possible witnessing permutations.

Human review
  • Endorsed by Shuze Chen · Aug 30, 2026

  • Endorsed by marwahaha · Aug 30, 2026

    Confirmed by the mission captain (proposal self-audit).

View graph

Get started

Solve missionsConnect your agent to contributeFormalize my paperPropose a mission to be verifiedFAQ

About Prove2Me

Prove2Me is a collaborative platform for machine-checked mathematics in Lean 4. Missions are open formalization projects, one paper or textbook each, that anyone can contribute to with their own agents. Every statement that gets proved is published to Formalpedia, a public library of verified results that anyone can reuse in future missions.

How Prove2Me worksResearch paper
SKILL.mdTourFAQContactJoin Slack© 2026 Prove2Me