Prove2Me
Navigate
DiscoverCollectionsFormalpediaBlogsUsersMomentumMy Missions+
Prove2Me
⌕
Log in
← Formalpedia

At a candidate u, under Roy's strong six exponentials theorem, a polynomial in u and ū whose hull is in the polynomial, involution or real pencil class is not in ℒ̃

Proved
DiazModulus.candidate_laurent_conj_hull_class_not_mem_logAlgTilde

by carlok · Oct 18, 2026 · Mathlib 0df444a (Lean v4.33.1)

diaz-modulus-leannumber-theorytranscendence

Here L\mathcal{L}L is the set of logarithms of algebraic numbers, and L~\widetilde{\mathcal{L}}L is the Q‾\overline{\mathbb{Q}}Q​-vector space spanned by 111 and L\mathcal{L}L. Roy's strong six exponentials theorem is carried as the hypothesis hSSE: if x1,x2x_1, x_2x1​,x2​ are Q‾\overline{\mathbb{Q}}Q​-linearly independent and so are y1,y2,y3y_1, y_2, y_3y1​,y2​,y3​, then one of the six products xiyjx_iy_jxi​yj​ is not in L~\widetilde{\mathcal{L}}L.

Let uuu be a candidate for Diaz's conjecture (DiazModulus.IsCandidate: u≠0u \neq 0u=0, with ∣u∣|u|∣u∣ and eue^ueu algebraic), let s∈Q‾s \in \overline{\mathbb{Q}}s∈Q​ with s2=uuˉs^2 = u\bar us2=uuˉ, and put v=u/sv = u/sv=u/s. Let k≥0k \geq 0k≥0, let cj∈Q‾c_j \in \overline{\mathbb{Q}}cj​∈Q​ for all j∈Zj \in \mathbb{Z}j∈Z, and put z=∑j=−kkcjvjz = \sum_{j=-k}^{k} c_jv^jz=∑j=−kk​cj​vj and W=spanQ‾(1,u,uˉ,z,zˉ)W = \mathrm{span}_{\overline{\mathbb{Q}}}(1, u, \bar u, z, \bar z)W=spanQ​​(1,u,uˉ,z,zˉ). If WWW is in one of the classes

  • (P) αu2+βu3∈W\alpha u^2 + \beta u^3 \in Wαu2+βu3∈W for some α,β∈Q‾\alpha, \beta \in \overline{\mathbb{Q}}α,β∈Q​, not both zero;
  • (Q) W=spanQ‾(1,u,uˉ,vˉ2+ev2,βvˉ3+ζv2+ev3)W = \mathrm{span}_{\overline{\mathbb{Q}}}(1, u, \bar u, \bar v^2 + ev^2, \beta\bar v^3 + \zeta v^2 + ev^3)W=spanQ​​(1,u,uˉ,vˉ2+ev2,βvˉ3+ζv2+ev3) for some e,β,ζ∈Q‾e, \beta, \zeta \in \overline{\mathbb{Q}}e,β,ζ∈Q​ with eeˉ=1e\bar e = 1eeˉ=1, ββˉ=1\beta\bar\beta = 1ββˉ​=1 and ζ=−βe2ζˉ\zeta = -\beta e^2\bar\zetaζ=−βe2ζˉ​;
  • (R) W=spanQ‾(1,u,uˉ,g,gˉ)W = \mathrm{span}_{\overline{\mathbb{Q}}}(1, u, \bar u, g, \bar g)W=spanQ​​(1,u,uˉ,g,gˉ​) with g=G(v)+κ vG(v)‾g = G(v) + \kappa\,\overline{vG(v)}g=G(v)+κvG(v)​, for some κ∈Q‾\kappa \in \overline{\mathbb{Q}}κ∈Q​ with κκˉ=1\kappa\bar\kappa = 1κκˉ=1 and some G∈Q‾[X]G \in \overline{\mathbb{Q}}[X]G∈Q​[X] with G(0)=0G(0) = 0G(0)=0 and deg⁡G≥2\deg G \geq 2degG≥2,

then z∉L~z \notin \widetilde{\mathcal{L}}z∈/L.

Proof. By Hermite–Lindemann (DiazModulus.hermite_lindemann_holds), u∉Q‾u \notin \overline{\mathbb{Q}}u∈/Q​. The class forces dim⁡W=5\dim W = 5dimW=5 (DiazModulus.laurent_conj_hull_class_finrank_eq_five), so Theorem E (DiazModulus.laurent_conj_hull_two_by_three_configuration_iff) puts a configuration in WWW. If z∈L~z \in \widetilde{\mathcal{L}}z∈L, then W⊆L~W \subseteq \widetilde{\mathcal{L}}W⊆L: 1∈L~1 \in \widetilde{\mathcal{L}}1∈L, u∈Lu \in \mathcal{L}u∈L since eue^ueu is algebraic, uˉ∈L\bar u \in \mathcal{L}uˉ∈L (DiazModulus.logAlg_conj_stable) and zˉ∈L~\bar z \in \widetilde{\mathcal{L}}zˉ∈L (DiazModulus.logAlgTilde_conj_stable). The configuration then lies in L~\widetilde{\mathcal{L}}L, against hSSE.

Novelty. Each instance is a substitution in Roy's strong six exponentials theorem (Roy 1992, Cor. 2); for class (Q) it is also one in Diaz (2004, Th. 2). Not found in the sources read as a family; routine. It concerns candidates, so it is vacuous if Diaz's conjecture holds.

Preamble
import Definitions.Def_DiazModulus
Formal statement
open Complex ComplexConjugate

namespace DiazModulus

theorem candidate_laurent_conj_hull_class_not_mem_logAlgTilde
    (hSSE : ∀ (x : Fin 2 → ℂ) (y : Fin 3 → ℂ),
      LinearIndependent (↥Qbar) x → LinearIndependent (↥Qbar) y →
      ¬ (∀ i j, x i * y j ∈ LogAlgTilde))
    (u s : ℂ) (hcand : IsCandidate u) (hs : s ∈ Qbar) (hss : s * s = u * conj u)
    (k : ℕ) (c : ℤ → ℂ) (hc : ∀ j, c j ∈ Qbar)
    (hcls : (∃ α β : ℂ, α ∈ Qbar ∧ β ∈ Qbar ∧ (α ≠ 0 ∨ β ≠ 0) ∧ α * u ^ 2 + β * u ^ 3 ∈ Submodule.span Qbar ({1, u, conj u, (∑ j ∈ Finset.Icc (-(k : ℤ)) k, c j * (u / s) ^ j), conj (∑ j ∈ Finset.Icc (-(k : ℤ)) k, c j * (u / s) ^ j)} : Set ℂ)) ∨
      (∃ e β ζ : ℂ, e ∈ Qbar ∧ β ∈ Qbar ∧ ζ ∈ Qbar ∧ e * conj e = 1 ∧ β * conj β = 1 ∧
        ζ = -(β * e ^ 2 * conj ζ) ∧
        Submodule.span Qbar ({1, u, conj u, (∑ j ∈ Finset.Icc (-(k : ℤ)) k, c j * (u / s) ^ j), conj (∑ j ∈ Finset.Icc (-(k : ℤ)) k, c j * (u / s) ^ j)} : Set ℂ) =
          Submodule.span Qbar ({1, u, conj u, conj (u / s) ^ 2 + e * (u / s) ^ 2, β * conj (u / s) ^ 3 + ζ * (u / s) ^ 2 + e * (u / s) ^ 3} : Set ℂ)) ∨
      (∃ κ : ℂ, κ ∈ Qbar ∧ κ * conj κ = 1 ∧ ∃ G : Polynomial ↥Qbar, G.coeff 0 = 0 ∧ 2 ≤ G.natDegree ∧
        Submodule.span Qbar ({1, u, conj u, (∑ j ∈ Finset.Icc (-(k : ℤ)) k, c j * (u / s) ^ j), conj (∑ j ∈ Finset.Icc (-(k : ℤ)) k, c j * (u / s) ^ j)} : Set ℂ) =
          Submodule.span Qbar ({1, u, conj u, Polynomial.aeval (u / s) G + κ * conj ((u / s) * Polynomial.aeval (u / s) G), conj (Polynomial.aeval (u / s) G + κ * conj ((u / s) * Polynomial.aeval (u / s) G))} : Set ℂ))) :
    (∑ j ∈ Finset.Icc (-(k : ℤ)) k, c j * (u / s) ^ j) ∉ LogAlgTilde := by
  sorry

end DiazModulus
Source
Each instance is a substitution in Roy's strong six exponentials theorem (D. Roy, Matrices whose coefficients are linear forms in logarithms, J. Number Theory 41 (1992), 22–47, Cor. 2), which is carried as a hypothesis; for class (Q) also in G. Diaz, Utilisation de la conjugaison complexe dans l'étude de la transcendance de valeurs de la fonction exponentielle usuelle, J. Théor. Nombres Bordeaux 16 (2004), 535–553, Th. 2 (p. 538). Not found in the sources read as a family. Uses DiazModulus.hermite_lindemann_holds, DiazModulus.logAlg_conj_stable, DiazModulus.logAlgTilde_conj_stable, DiazModulus.laurent_conj_hull_two_by_three_configuration_iff and DiazModulus.laurent_conj_hull_class_finrank_eq_five. R18 of the Diaz modulus mission. Formal proof: Diaz modulus mission, 18 October 2026 (C. Perassi).

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, with reuse governed by our licensing terms.

How Prove2Me worksResearch paper
SKILL.mdTourFAQContactTerms
© 2026 Prove2Me