Binary and i.i.d. tensor additivity
ProvedQAlgorithms.Papers.Independent.Support50.P23.alphaAdditivityLet and reference states be normalized on arbitrary finite-dimensional nonzero complex systems. For every , with factors regrouped as ,
In particular, for every integer ,
The two factors can have different dimensions. These identities are the product scaling used in the finite-block bound.
import Definitions.Def_QAlgorithms_Papers_Independent_Support50_P23
noncomputable section
open Filter
open scoped ENNReal ComplexConjugate ComplexOrder
namespace QAlgorithms.Papers.Independent.Support50.P23
attribute [local instance] Classical.propDecidable
-- Exact accepted bodies. Other similarly named catalogue entries use different
-- smoothing, conditioning order, logarithm base, or channel-level objects.
open QAlgorithms.Papers.Independent.Gap20.P03 (fixedMin)
open QAlgorithms.Papers.Independent.Dependency50.P36 (relativeEntropy)
open QAlgorithms.Papers.Independent.Information100.P67 (IsSubnormalized)
open QAlgorithms.Papers.Independent.Information100.P69 (matrixPower)
variable {A B C D I : Type*}
variable [Fintype A] [Fintype B] [Fintype C] [Fintype D] [Fintype I]
variable [DecidableEq A] [DecidableEq B] [DecidableEq C] [DecidableEq D] [DecidableEq I]
/-- Section II, PDF p.4, (18): arbitrary binary tensor additivity and its displayed
i.i.d. specialization for positive n, including all endpoint orders. -/
theorem alphaAdditivity [Nonempty A] [Nonempty B]
(rho : Matrix (A × B) (A × B) ℂ) (sigma : Matrix B B ℂ)
(hrho : IsDensityMatrix rho) (hsigma : IsDensityMatrix sigma) (alpha : ℝ≥0∞) :
(∀ {C D : Type*} [Fintype C] [Fintype D] [DecidableEq C] [DecidableEq D]
[Nonempty C] [Nonempty D]
(tau : Matrix (C × D) (C × D) ℂ) (omega : Matrix D D ℂ),
IsDensityMatrix tau → IsDensityMatrix omega →
alphaEntropy (independentProduct rho tau) (Matrix.kronecker sigma omega) alpha =
alphaEntropy rho sigma alpha + alphaEntropy tau omega alpha) ∧
(∀ n : ℕ, 0 < n →
alphaEntropy (jointPower rho n) (matrixPower sigma n) alpha =
n • alphaEntropy rho sigma alpha) := by sorry
end QAlgorithms.Papers.Independent.Support50.P23
end
Read-back
What the Lean code literally says, in plain math · gpt-6.1-sol
Let be nonempty finite types with decidable equality, let on and on be positive semidefinite matrices of trace , and let . For any such pair , write , set if , and set it to otherwise; define at , when and that support inclusion fails, and otherwise , where assigns at zero eigenvalues and , for . Define for finite positive , , and , with extended-real arithmetic. Then for every pair of nonempty finite types with decidable equality and every pair of positive semidefinite trace-one matrices on and on ,
where . Also, for every natural , , where tensor-power entries are coordinatewise products and denotes repeated extended-real addition. The order endpoints are included, but the tensor-power assertion excludes .
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.