Every model has finite entire exponential preimage equations
ProvedWeierstrassEllipticZeta.exponentialPreimage_finite_entire_equationsalgebraic-geometryalgebraic-groupselliptic-curves
For every compatible Model M and actual algebraic subgroup H, pull back all bihomogeneous equations vanishing on H under the explicit entire exponential coordinates. Their common zero locus in C³ is defined by finitely many entire functions, by the Hilbert basis theorem. This result does not assert that the inverse image is additive and does not use connectedness or a chosen canonical group model.
Preamble
import Definitions.Def_WeierstrassEllipticZeta_ExponentialPreimage set_option autoImplicit false set_option backward.isDefEq.respectTransparency false open WeierstrassEllipticZeta WeierstrassEllipticZeta.PhilipponApplication PhilipponMultiplicity
Formal statement
theorem WeierstrassEllipticZeta.exponentialPreimage_finite_entire_equations
(S : Fin 5 → ℂ → ℂ) (hS : ∀ j, AnalyticOnNhd ℂ (S j) Set.univ)
(M : Model S) (H : AlgebraicSubgroup M.group) :
∃ F : Finset ((Fin 3 → ℂ) → ℂ),
(∀ f ∈ F, AnalyticOnNhd ℂ f Set.univ) ∧
∀ v, v ∈ M.exponentialPreimage H ↔ ∀ f ∈ F, f v = 0 := by sorrySource
Senthil Kumar, https://doi.org/10.1017/S001309152610145X, Appendix A, exponential maps and Lemma A.1. Supporting application geometry belongs to Senthil, not Philippon.