AFTD Auto-Formalizing Theoretical Domains

matrix3_trace_saturation

theoremverifiedformalization

If a cyclic trace decomposition over a field has first-factor matrix ranks summing to 27, then a_t * b_s * c_t * a_s equals a_s for t=s and zero otherwise.

Statement

theorem matrix3_trace_saturation {K : Type*} [Field K] {r : ℕ} (a b c : Fin r → Matrix (Fin 3) (Fin 3) K)
    (h : ∀ i j j' k k' i' : Fin 3,
      ∑ s, a s i j * b s j' k * c s k' i' =
        if j = j' ∧ k = k' ∧ i = i' then 1 else 0)
    (hd : ∑ s, (a s).rank = 27) :
    ∀ t s, a t * b s * c t * a s = if t = s then a s else 0
Formalizes
arXiv:2610.11823v3, Section 4.2, equation (4.2) with excess zero. The proof here uses the injectivity of the split factor map at equal dimensions and applies over every field; no computational certificates are assumed.
Verified
29 Oct 2026
Axioms
Classical.choiceQuot.soundpropext
Built on 1
Used by 1

The proof, explained

A cyclic trace decomposition gives a linear map whose dimension matches that of the target space when matrix ranks sum to 27. Since the map is surjective by reconstruction, it is injective, making the reconstruction map a two-sided inverse. Evaluating this two-sided inverse on a basis column isolates the components, showing the saturated product is a_s when t = s and 0 otherwise.

  1. Linearity of matrix3_saturation_map with respect to addition

    By extensionality on p, expanding the sum and distributing multiplication over addition.

    lines 11–13

    uses add_mul, Finset.sum_add_distrib

      map_add' x y := by
        ext p
        simp [add_mul, Finset.sum_add_distrib]
  2. Linearity of matrix3_saturation_map with respect to scalar multiplication

    By extensionality on p, factoring the scalar out of the sum.

    lines 14–16

    uses Finset.mul_sum, mul_assoc

      map_smul' c x := by
        ext p
        simp [Finset.mul_sum, mul_assoc]
  3. Right inverse reconstruction property

    By extensionality on p, using the reconstruction theorem for cyclic trace decompositions.

    lines 28–30

    uses matrix3_trace_split_reconstruction

        matrix3_saturation_map a b (matrix3_saturation_preimage a c z) = z := by
      ext p
      exact matrix3_trace_split_reconstruction a b c h z p
  4. Main theorem test

    Main proof.

    lines 39–73

      classical
      have hdim : Module.finrank K (matrix3_saturation_space a) =
          Module.finrank K (Fin 3 × Fin 3 × Fin 3 → K) := by
        simpa [matrix3_saturation_space, Module.finrank_pi, Module.finrank_pi_fintype, Matrix.rank] using hd
      have hinj : Function.Injective (matrix3_saturation_map a b) :=
        (LinearMap.injective_iff_surjective_of_finrank_eq_finrank hdim).mpr
          (fun z => ⟨matrix3_saturation_preimage a c z, matrix3_saturation_reconstruction a b c h z⟩)
      have hyx (x : matrix3_saturation_space a) : matrix3_saturation_preimage a c (matrix3_saturation_map a b x) = x := by
        apply hinj
        exact matrix3_saturation_reconstruction a b c h _
      intro t s
      ext i j
      let v : LinearMap.range (a s).mulVecLin :=
        ⟨(a s).col j, ⟨Pi.single j 1, by simp⟩⟩
      have hv := congrArg (fun x : matrix3_saturation_space a => (x t).val i) (hyx (Pi.single s v))
      have hsimp : ∀ p, (matrix3_saturation_map a b (Pi.single s v)) p =
          a s p.1 j * b s p.2.1 p.2.2 := by
        intro p
        change (∑ q, ((Pi.single s v : matrix3_saturation_space a) q).val p.1 * b q p.2.1 p.2.2) = _
        rw [Finset.sum_eq_single s]
        · simp [v]
        · intro q _ hqs
          simp [Pi.single_eq_of_ne hqs]
        · simp
      simp only [matrix3_saturation_preimage, mulVecLin_apply,
        mulVec, dotProduct, hsimp] at hv
      have hexpand : (a t * b s * c t * a s) i j =
          ∑ q, a t i q * (∑ k, ∑ i', c t k i' * (a s i' j * b s q k)) := by
        simp only [Matrix.mul_apply, Fin.sum_univ_three]
        ring
      rw [hexpand, hv]
      by_cases hts : t = s
      · subst t
        simp [v]
      · simp [hts, v]

Written by a model (gemini-3.8-flash) from the Lean proof and the goal states Lean reported at each step. Lean verified the proof; this explanation is a reading aid and is not itself machine-checked.

Lean source view module on GitHub

open Matrix Finsetnoncomputable sectionvariable {K : Type*} [Field K] {r : ℕ}private abbrev matrix3_saturation_space (a : Fin r → Matrix (Fin 3) (Fin 3) K) :=  ∀ s, LinearMap.range (a s).mulVecLinprivate def matrix3_saturation_map (a b : Fin r → Matrix (Fin 3) (Fin 3) K) :    matrix3_saturation_space a →ₗ[K] (Fin 3 × Fin 3 × Fin 3 → K) where  toFun x p := ∑ s, (x s).val p.1 * b s p.2.1 p.2.2  map_add' x y := by    ext p    simp [add_mul, Finset.sum_add_distrib]  map_smul' c x := by    ext p    simp [Finset.mul_sum, mul_assoc]private def matrix3_saturation_preimage (a c : Fin r → Matrix (Fin 3) (Fin 3) K)    (z : Fin 3 × Fin 3 × Fin 3 → K) : matrix3_saturation_space a :=  fun s => ⟨(a s).mulVecLin (fun j => ∑ k, ∑ i, c s k i * z (i,j,k)),    ⟨_, rfl⟩⟩private theorem matrix3_saturation_reconstruction (a b c : Fin r → Matrix (Fin 3) (Fin 3) K)    (h : ∀ i j j' k k' i' : Fin 3,      ∑ s, a s i j * b s j' k * c s k' i' =        if j = j' ∧ k = k' ∧ i = i' then 1 else 0)    (z : Fin 3 × Fin 3 × Fin 3 → K) :    matrix3_saturation_map a b (matrix3_saturation_preimage a c z) = z := by  ext p  exact matrix3_trace_split_reconstruction a b c h z pendtheorem matrix3_trace_saturation {K : Type*} [Field K] {r : ℕ} (a b c : Fin r → Matrix (Fin 3) (Fin 3) K)    (h : ∀ i j j' k k' i' : Fin 3,      ∑ s, a s i j * b s j' k * c s k' i' =        if j = j' ∧ k = k' ∧ i = i' then 1 else 0)    (hd : ∑ s, (a s).rank = 27) :    ∀ t s, a t * b s * c t * a s = if t = s then a s else 0 := by  classical  have hdim : Module.finrank K (matrix3_saturation_space a) =      Module.finrank K (Fin 3 × Fin 3 × Fin 3 → K) := by    simpa [matrix3_saturation_space, Module.finrank_pi, Module.finrank_pi_fintype, Matrix.rank] using hd  have hinj : Function.Injective (matrix3_saturation_map a b) :=    (LinearMap.injective_iff_surjective_of_finrank_eq_finrank hdim).mpr      (fun z => ⟨matrix3_saturation_preimage a c z, matrix3_saturation_reconstruction a b c h z⟩)  have hyx (x : matrix3_saturation_space a) : matrix3_saturation_preimage a c (matrix3_saturation_map a b x) = x := by    apply hinj    exact matrix3_saturation_reconstruction a b c h _  intro t s  ext i j  let v : LinearMap.range (a s).mulVecLin :=    ⟨(a s).col j, ⟨Pi.single j 1, by simp⟩⟩  have hv := congrArg (fun x : matrix3_saturation_space a => (x t).val i) (hyx (Pi.single s v))  have hsimp : ∀ p, (matrix3_saturation_map a b (Pi.single s v)) p =      a s p.1 j * b s p.2.1 p.2.2 := by    intro p    change (∑ q, ((Pi.single s v : matrix3_saturation_space a) q).val p.1 * b q p.2.1 p.2.2) = _    rw [Finset.sum_eq_single s]    · simp [v]    · intro q _ hqs      simp [Pi.single_eq_of_ne hqs]    · simp  simp only [matrix3_saturation_preimage, mulVecLin_apply,    mulVec, dotProduct, hsimp] at hv  have hexpand : (a t * b s * c t * a s) i j =      ∑ q, a t i q * (∑ k, ∑ i', c t k i' * (a s i' j * b s q k)) := by    simp only [Matrix.mul_apply, Fin.sum_univ_three]    ring  rw [hexpand, hv]  by_cases hts : t = s  · subst t    simp [v]  · simp [hts, v]
See it in the graphDiscuss this resultChallenge it