K24 §4: conditional Neumann transfer with inverse bound
ProvedSiweiK24.neumann_transferLet be real Banach spaces, a bounded linear isomorphism with bounded inverse, and a bounded linear operator. For a real constant , assume
Then there exists a bounded linear isomorphism with
This formalizes only the abstract inverse-transfer step in K24 §4, equations (4.1)–(4.3). The source already supplies its informal Neumann-series argument; it is a known conditional lemma awaiting a Lean proof, not a new unsolved research conjecture. It does not assert K24's geometric change-of-Bianchi estimate, the missing same-window organizer, or the four-manifold conjecture. In particular, the perturbation norm hypothesis must be established separately for any proposed geometric application.
Formalization Note. This exact supporting statement compiled successfully in the pinned environment on 23 October 2026. Its body is an unproved placeholder; it does not represent the full geometric K24 target.
import Mathlib.Analysis.Normed.Operator.NormedSpace set_option autoImplicit false
namespace SiweiK24
/-- K24, Section 4, equations (4.1)--(4.3): the abstract Neumann-transfer step.
This is a statement awaiting a Lean proof, not a claim that the geometric
same-window hypotheses or the Einstein conjecture have been established. -/
theorem neumann_transfer
(X Y : Type*)
[NormedAddCommGroup X] [NormedSpace ℝ X] [CompleteSpace X]
[NormedAddCommGroup Y] [NormedSpace ℝ Y] [CompleteSpace Y]
(A : X ≃L[ℝ] Y) (R : X →L[ℝ] Y) (C_A : ℝ)
(hC : 0 < C_A)
(hA : ‖A.symm.toContinuousLinearMap‖ ≤ C_A)
(hR : C_A * ‖R‖ < (1 / 2 : ℝ)) :
∃ B : X ≃L[ℝ] Y,
B.toContinuousLinearMap = A.toContinuousLinearMap + R ∧
‖B.symm.toContinuousLinearMap‖ ≤ 2 * C_A := by sorry
end SiweiK24Read-back
What the Lean code literally says, in plain math · GPT-6 (independent blind Codex sub-agent; exact variant not exposed)
For every pair of complete real normed vector spaces and , every continuous real-linear equivalence (a bijective continuous real-linear map with continuous inverse), every continuous real-linear map , and every real number , if , the operator norm of the inverse satisfies , and , then there exists a continuous real-linear equivalence whose underlying continuous linear map is exactly the sum of the underlying map of and , meaning for every , and whose inverse has operator norm . All norms of maps in this assertion are operator norms induced by the given norms on and ; both spaces have a normed additive commutative group structure, a compatible real vector-space structure, and are complete. The perturbation bound is strict, whereas both inverse-norm bounds are non-strict; the conclusion asserts existence and does not include a separate uniqueness assertion. No nonzero-dimension assumption is made: the zero vector space is allowed, and if one of or is the zero space, existence of forces the other to be the zero space as well, in which case the map norms are zero and the hypotheses hold for every . The case is also included, with the perturbation hypothesis then automatic; and negative are excluded by the positivity hypothesis.
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.