AFTD Auto-Formalizing Theoretical Domains

linear_flattening_testable_zero_order

theoremverified

For every n ≥ 1, σ_0((ℂⁿ)⊗³) is testable by the linear flattening method (some linear flattening has potency greater than 0), so the predicate linear_flattening_testable is not vacuously false.

Statement

theorem linear_flattening_testable_zero_order (n : ℕ) (hn : 1 ≤ n) :
    linear_flattening_testable n 0
Source
step towards linear_flattening_testable_even_step
Verified
23 Oct 2026
Axioms
Classical.choiceQuot.soundpropext
Built on 3
  • linear_flattening_testableThe secant variety σ_r((ℂⁿ)⊗³) is testable by the linear flattening method if there exists a linear flattening L : (ℂⁿ)⊗³ → Mat_{u×v}(ℂ)…
  • tensor3_pureThe pure tensor x ⊗ y ⊗ z in ℂⁿ ⊗ ℂⁿ ⊗ ℂⁿ, with entries x_i y_j z_k.
  • linear_flattening_potency_gtA linear flattening L : ℂⁿ ⊗ ℂⁿ ⊗ ℂⁿ → Mat_{u×v}(ℂ) has potency greater than r: some tensor T has rank L(T) > r · ρ(L), where ρ(L) =…
Used by 1

The proof, explained

The theorem establishes that the order-0 linear flattening testability condition holds for every dimension n ≥ 1. We construct a linear map L from 3-tensors to 1 × 1 matrices by evaluating the tensor at the index (0, 0, 0). Evaluating L on the all-ones tensor yields the 1 × 1 identity matrix, which has rank 1; for order r = 0, the testability condition reduces to 0 < 1.

  1. Construct a linear map L from 3-tensors to 1 × 1 matrices by extracting the (0, 0, 0) entry.

    Since n ≥ 1, the index 0 is valid in Fin n. We define L to map any tensor A to the 1 × 1 matrix whose single entry is A(0, 0, 0). Linearity under addition and scalar multiplication holds definitionally from pointwise operations on functions.

    lines 4–7

      let L : (Fin n → Fin n → Fin n → ℂ) →ₗ[ℂ] Matrix (Fin 1) (Fin 1) ℂ :=
        { toFun := fun A => fun _ _ => A ⟨0, hn⟩ ⟨0, hn⟩ ⟨0, hn⟩,
          map_add' := by intro a b; ext i j; rfl,
          map_smul' := by intro c a; ext i j; rfl }
  2. Witness the existential quantifiers with dimensions 1 × 1, linear map L, and the all-ones tensor T, and introduce arbitrary vectors x, y, z.

    To establish linear_flattening_testable n 0, we provide dimensions u = 1 and v = 1, the linear flattening L, and the tensor T identically equal to 1. For arbitrary vectors x, y, z ∈ ℂⁿ defining a pure tensor, this leaves the potency condition 0 · rank(L(x ⊗ y ⊗ z)) < rank(L(T)).

    lines 8–9

      refine ⟨1, 1, L, (fun _ _ _ => (1 : ℂ)), ?_⟩
      intro x y z
  3. The image L(T) of the all-ones tensor is the 1 × 1 identity matrix.

    By extensionality, we compare entry (i, j) on both sides. The (i, j) entry of the identity matrix is 1 if i = j and 0 otherwise (Matrix.one_apply). Because Fin 1 is a subsingleton, any two indices are equal (Subsingleton.elim), so the identity entry is 1, matching L(T) i j = T(0, 0, 0) = 1.

    lines 10–14

    uses Matrix.one_apply, Subsingleton.elim

      have hLT : L (fun _ _ _ => (1 : ℂ)) = 1 := by
        ext i j
        rw [Matrix.one_apply]
        simp only [Subsingleton.elim i j, ↓reduceIte]
        rfl
  4. The potency inequality 0 < 1 holds, completing the proof.

    Substituting L(T) = 1, the rank of the 1 × 1 identity matrix is the cardinality of Fin 1, which is 1 (Matrix.rank_one). The condition reduces to 0 · rank(L(x ⊗ y ⊗ z)) < 1, or simply 0 < 1, which simplification closes immediately.

    lines 15–16

    uses Matrix.rank_one

      rw [hLT, Matrix.rank_one]
      simp
Matrix.one_apply
The (i, j)-th entry of the identity matrix is 1 if i = j and 0 otherwise.
Subsingleton.elim
Any two elements of a subsingleton type are equal.
Matrix.rank_one
The rank of the identity matrix is the cardinality of its indexing type.

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

/-- Non-vacuity of the testability predicate: for every n ≥ 1 the order-0 case holds. -/theorem linear_flattening_testable_zero_order (n : ℕ) (hn : 1 ≤ n) :    linear_flattening_testable n 0 := by  let L : (Fin n → Fin n → Fin n → ℂ) →ₗ[ℂ] Matrix (Fin 1) (Fin 1) ℂ :=    { toFun := fun A => fun _ _ => A ⟨0, hn⟩ ⟨0, hn⟩ ⟨0, hn⟩,      map_add' := by intro a b; ext i j; rfl,      map_smul' := by intro c a; ext i j; rfl }  refine ⟨1, 1, L, (fun _ _ _ => (1 : ℂ)), ?_⟩  intro x y z  have hLT : L (fun _ _ _ => (1 : ℂ)) = 1 := by    ext i j    rw [Matrix.one_apply]    simp only [Subsingleton.elim i j, ↓reduceIte]    rfl  rw [hLT, Matrix.rank_one]  simp
See it in the graphDiscuss this resultChallenge it