Lemma 8
ProvedQAlgorithms.Papers.Independent.Support50.P21.minRelative_joint_convexityAll systems are nonzero finite-dimensional complex Hilbert spaces, and logarithms have base two.
Let be a finite nonempty index set, let with , and let be states and with . Then
Zero weights are allowed; the support conditions make every summand finite. This is joint convexity on the proof's stated domain.
import Definitions.Def_QAlgorithms_Papers_Independent_Support50_P21
noncomputable section
namespace QAlgorithms.Papers.Independent.Support50.P21
open QAlgorithms QAlgorithms.Entanglement Filter
open scoped Matrix ComplexConjugate ComplexOrder
attribute [local instance] Classical.propDecidable Classical.decEq
-- Use the accepted declarations directly; no paper-local aliases hide dimension binders.
open QAlgorithms.Papers.Independent.Claimed50.P33
(FiniteState IsAverageLocalMonotone IsAdaptiveLOCCMonotone)
open QAlgorithms.Papers.Independent.Information100.P63 (Mat)
open QAlgorithms.Papers.Independent.Dependency50.P35 (tr elog maxRelative)
open QAlgorithms.Papers.Independent.Information100.P67 (IsSubnormalized)
open QAlgorithms.Papers.Independent.Information100.P07 (support)
open QAlgorithms.Papers.Independent.Information100.P71 (supportPower SupportIncluded)
variable {I J : Type} [Fintype I] [DecidableEq I] [Fintype J] [DecidableEq J]
section Statements
variable [Nonempty I] [Nonempty J]
/-- Lemma 8, PDF pp.4–5, (24): joint convexity on the precise proof domain.
Zero weights are allowed; support inclusion makes each summand finite. -/
theorem minRelative_joint_convexity {K : Type} [Fintype K] [Nonempty K]
(p : K → ℝ) (hp : ∀ i, 0 ≤ p i) (hsum : ∑ i, p i = 1)
(rho sigma : K → Mat I) (hrho : ∀ i, IsDensityMatrix (rho i))
(hsigma : ∀ i, (sigma i).PosSemidef) (hsupp : ∀ i, SupportIncluded (rho i) (sigma i)) :
dMin (∑ i, (p i : ℂ) • rho i) (∑ i, (p i : ℂ) • sigma i) ≤
∑ i, (p i : EReal) * dMin (rho i) (sigma i) := by sorry
end Statements
end QAlgorithms.Papers.Independent.Support50.P21
end
Read-back
What the Lean code literally says, in plain math · gpt-6.1-sol
For any finite index type with decidable equality and any finite nonempty type , let for , with . Let be complex matrices on such that every is positive semidefinite with complex trace one, every is positive semidefinite, and for every . Let be the orthogonal projection onto the range of , write , and define the extended-real quantity , where for and otherwise. Then
where the right side is an extended-real finite sum. Zero weights are allowed, and the positivity and support hypotheses are required even at indices with zero weight; the matrices need not have trace one.
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.