eq44
OpenQAlgorithms.Papers.Independent.Support50.P15.eq44Let , be Hermitian, , and a normalized full-rank state. Put , , and let be available work, identified with the process supremum in Eqs. (41)--(43). For every ,
For a fixed local dimension and temperature , consider any sequence of such systems with and entropy-density changes . Then
The liminf is in the extended real order, so unbounded work sequences are retained. The printed asymptotic sign does not supply the coefficient-two bound in every finite dimension.
import Definitions.Def_QAlgorithms_Papers_Independent_Support50_P15
noncomputable section
attribute [local instance] Classical.propDecidable
open scoped Matrix ComplexConjugate ENNReal Matrix.Norms.L2Operator
open Filter MeasureTheory
namespace QAlgorithms.Papers.Independent.Support50.P15
open QAlgorithms.Papers.Independent.Information100.P56 (entropy)
/-- Eq. (44), PDF p.13. The printed asymptotic sign is not a finite-d inequality with coefficient 2.
The liminf takes place in EReal, retaining unbounded work sequences. -/
theorem eq44 :
(∀ d : ℕ, 2 ≤ d → ∀ h : Matrix (Fin d) (Fin d) ℂ, h.IsHermitian →
∀ t : ℝ, 0 < t → ∀ a : EntropyState d, fullRank a.val →
(t : EReal) * tightBound d (-difference (thermalFinite h t) a.val) ≤ (availableWork h t a : EReal) ∧
(∀ n : ℝ, maxVariance d ≤ n →
((t * exponentialBound n (-difference (thermalFinite h t) a.val) : ℝ) : EReal) ≤
(t : EReal) * tightBound d (-difference (thermalFinite h t) a.val))) ∧
(∀ (l : ℕ), 2 ≤ l → ∀ t : ℝ, 0 < t →
∀ (h : (n : ℕ) → Matrix (Fin (l ^ n)) (Fin (l ^ n)) ℂ)
(a : (n : ℕ) → EntropyState (l ^ n)),
(∀ n, (h n).IsHermitian ∧ fullRank (a n).val) →
∀ s : ℝ, Tendsto (fun n : ℕ => difference (thermalFinite (h n) t) (a n).val / n) atTop (nhds s) →
((2 * t * (s / Real.log l) ^ 2 : ℝ) : EReal) ≤
Filter.liminf (fun n => (availableWork (h n) t (a n) : EReal)) atTop) := by sorry
end QAlgorithms.Papers.Independent.Support50.P15
end
Read-back
What the Lean code literally says, in plain math · gpt-6.1-sol
Use natural logarithms and write for the entropy of a density matrix, with . For , put , , and . Define the extended real binary relative entropy , where if , if and , and otherwise . Set
where an empty extended-real infimum is . For a Hermitian and positive , let and
The declaration asserts two statements. First, for every natural , every Hermitian complex matrix , every real , and every density matrix with zero kernel, writing , one has
as extended-real inequalities. Second, for every natural , every fixed real , and sequences indexed by all natural numbers, where is a Hermitian complex matrix and is a density matrix of that dimension with zero kernel, if a real satisfies
then
in the extended reals. The sequences include , where their dimension is and the displayed quotient uses the convention that real division by zero is zero.
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.