Convergence of every separable input to an equal-block fixed point
OpenQAlgorithms.Papers.Independent.Intakenext100.P68.separableFixedPointLet be finite local mode counts, with interleaved quadratures , , and doubled covariance . Thus vacuum covariance is . Put . A covariance matrix (CM) is real symmetric and satisfies . All inequalities below are Loewner inequalities.
Let be a physical CM. Construct the sequence as follows: a nonphysical matrix is followed by zero; otherwise put , , and . All inverses are spectral Moore–Penrose inverses. Every positive-index matrix has split . Set for , where the norm is Euclidean induced. Separability always means the trace-norm closed convex hull of full oscillator product densities, independently of covariance witness inequalities.
If is separable as an actual Gaussian covariance, then there is an -mode CM such that
where is the source's one-step covariance map. Convergence is in the usual finite-matrix topology. There is no strict margin, convergence rate or finite stopping hypothesis. Only the positive-index tail is placed in a single fixed-size space.
import Definitions.Def_QAlgorithms_Papers_Independent_Intakenext100_P68
noncomputable section
open MeasureTheory Filter
open scoped BigOperators Topology ComplexOrder
namespace QAlgorithms.Papers.Independent.Intakenext100.P68
attribute [local instance] Classical.propDecidable Classical.decEq
open QAlgorithms.Papers.Independent.Dependency50.P33
(State PureState IsSeparable IsBarycenter productCoeff)
open QAlgorithms.Papers.Independent.Dependency50.P10 (complexify)
open QAlgorithms.Papers.Independent.Frontier30.P08 (QuadratureIndex)
open QAlgorithms.Papers.Independent.Claimed50.P19 (Cov)
open QAlgorithms.Papers.Independent.Claimed50.P47 (Occupation)
/-- PDF p.3, bottom-left paragraph continued top-right: all separable inputs
converge to diag(Ainf,Ainf), a physical fixed point. No strict margin, rate or
finite stopping is assumed. Convergence uses only the fixed-size positive tail. -/
theorem separableFixedPoint {n m : ℕ} (hn : 0 < n) (hm : 0 < m)
(g : CovarianceBlocks n m) (hg : IsPhysicalCovariance g) (hsep : SeparableCM g) :
∃ U : Cov n, IsCM U ∧
Tendsto (fun k : ℕ => covarianceBlockMatrix (tail g k)) atTop
(nhds (realCovarianceSum U U)) ∧
covarianceStep ({ A := U, B := U, C := 0 } : CovarianceBlocks n n) =
({ A := U, B := U, C := 0 } : CovarianceBlocks n n) := by sorry
end QAlgorithms.Papers.Independent.Intakenext100.P68
end
Read-back
What the Lean code literally says, in plain math · gpt-6.1-sol
Let be natural numbers and a real block triple of sizes . Write , , and assume is symmetric and , using entrywise complexification. Assume there exist a density state on and a real interleaved vector with the following properties. Here density means that the coefficients admit a nuclear expansion with and , every finite quadratic form is real and nonnegative, and the real diagonal is summable with sum . Let be the infimum in of these nuclear expansion sums, with empty infimum . For every , assume some finite convex combination of product density matrices satisfies . Define
and let be the product over both subsystems’ modes of . For every real interleaved vector , assume that each inner sum and the outer sum in are summable and have value , where . Define as Hermitian spectral functional calculus with , taking , or zero for non-Hermitian . For with modes, put ; define if is symmetric and , and zero in all blocks otherwise. Set , . Then there exists a real symmetric matrix with such that
The convergence is in the usual finite-dimensional topology on real matrices.
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.