Convexity of total displacement Fisher information, Eq. (15)
OpenQAlgorithms.Papers.Independent.Intakenext100.P62.fisherConvexityClosedLet and let be arbitrary normalized states on the full -mode bosonic Fock spaces. Their joint input is the product ; entanglement within either input is unrestricted. Let retain the port with quadratures . Write for extended nonnegative von Neumann entropy in nats.
For each state, let be the sum of its divergence-based displacement Fisher information over all canonical directions, without averaging. Require local finiteness and genuine second-derivative regularity for every directional divergence of both inputs and their output. Then, for all ,
The same displacement scale is used throughout. This is convexity under beam-splitter mixing on the finite Fisher domain.
import Definitions.Def_QAlgorithms_Papers_Independent_Intakenext100_P62
noncomputable section
namespace QAlgorithms.Papers.Independent.Intakenext100.P62
open scoped ENNReal BigOperators
open QAlgorithms.Papers.Independent.Dependency50.P33 (State productCoeff)
open QAlgorithms.Papers.Independent.Frontier18.P07 (extendedEntropyNats)
open QAlgorithms.Papers.Independent.Claimed50.P19 (FockIndex)
open QAlgorithms.Papers.Independent.Claimed50.P39 (g displacementNoiseChannel)
open QAlgorithms.Papers.Independent.Frontier18.P18 (operatorCoeff boundedOperator)
open QAlgorithms.Papers.Independent.Gap48.P26 (weightedPhotonExpectation)
open QAlgorithms.Papers.Independent.Bridge50.P28
(classicalPhotonCapacityBits thermalNoiseChannelKernel coherentSignalRateBits)
open QAlgorithms.Papers.Independent.Next46.P17
(beamSplitter bosonicEntropyPower diffusionChannel TotalFisherRegular totalFisher)
/-- PDF p.5, (15): total displacement Fisher convexity under beam-splitter
mixing, including lambda=0,1. Every one of the 2n directional second derivatives
is in its genuine locally finite differentiability domain for both inputs and
output. A common displacement or bits/nats rescaling cancels from this inequality. -/
theorem fisherConvexityClosed {n : ℕ} (hn : 0 < n)
(lam : ℝ) (hlam0 : 0 ≤ lam) (hlam1 : lam ≤ 1)
(rhoX rhoY : State (FockIndex n))
(hX : TotalFisherRegular rhoX.coefficients)
(hY : TotalFisherRegular rhoY.coefficients)
(hOut : TotalFisherRegular (beamSplitter lam
(productCoeff rhoX.coefficients rhoY.coefficients))) :
totalFisher (beamSplitter lam (productCoeff rhoX.coefficients rhoY.coefficients)) ≤
lam * totalFisher rhoX.coefficients +
(1 - lam) * totalFisher rhoY.coefficients := by sorry
end QAlgorithms.Papers.Independent.Intakenext100.P62
end
Read-back
What the Lean code literally says, in plain math · gpt-6.1-sol
For every positive integer , every real , and every pair of states on , assume the regularity condition defined below holds for each of , , and . The theorem asserts
A state admits an expansion with and , has real nonnegative finite quadratic forms, and has summable real diagonal of sum . The output matrix is
where
For , define
where
The orbit of any matrix is
For matrices on , define
Here the kernels are those of finite-matrix multiplication on complex coordinate vectors, embedded in , uses continuous functional calculus of the total real logarithm, and the supremum includes the empty submatrix. Set
where , and the second derivative means the derivative of the total derivative function. The required regularity, separately for every and for each , is that on some neighborhood of the displayed relative entropy is finite, on some neighborhood of the real function is differentiable at every point, and the function is differentiable at . These neighborhoods need not be the same. The assertion includes , excludes , and imposes no separate energy assumption; all complex infinite sums use their stated grouping and the total nonsummable-family value .
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.