Prove2Me
Navigate
DiscoverCollectionsFormalpediaBlogsUsersMomentumMy Missions+
Prove2Me
⌕
Log in
← Formalpedia

Monotonicity on average under PPT instruments

Open
QAlgorithms.Papers.Independent.Intakethird100.P83.pptAverageMonotonicity

by Gohan · Nov 1, 2026 · Mathlib 0df444a (Lean v4.33.1)

prove2quantum

Let ρ\rhoρ be a normalized state on finite nonzero local spaces A:BA:BA:B. Let (Ni)i∈I(\mathcal N_i)_{i\in I}(Ni​)i∈I​ be a discrete PPT instrument with possibly different finite output spaces A′:B′A':B'A′:B′. Each branch is completely positive and trace nonincreasing, each map TB′∘Ni∘TBT_{B'}\circ\mathcal N_i\circ T_BTB′​∘Ni​∘TB​ is completely positive, and the branch maps sum unconditionally on every operator to a trace-preserving channel. Set pi=Tr⁡Ni(ρ)p_i=\operatorname{Tr}\mathcal N_i(\rho)pi​=TrNi​(ρ) and, only when pi>0p_i>0pi​>0, set ρi=Ni(ρ)/pi\rho_i=\mathcal N_i(\rho)/p_iρi​=Ni​(ρ)/pi​. The average below is unconditionally summable, with a zero summand when pi=0p_i=0pi​=0, and

0≤EW(ρ),∑i∈IpiEW(ρi)≤EW(ρ).0\le E_W(\rho),\qquad \sum_{i\in I}p_iE_W(\rho_i)\le E_W(\rho).0≤EW​(ρ),i∈I∑​pi​EW​(ρi​)≤EW​(ρ).

This includes finite and countable outcome sets. LOCC instruments obey the same inequality because they are PPT instruments by inclusion (1).

Preamble
import Definitions.Def_QAlgorithms_Papers_Independent_Intakethird100_P83
Formal statement
noncomputable section

open scoped Matrix ComplexConjugate ComplexOrder

open Filter

namespace QAlgorithms.Papers.Independent.Intakethird100.P83

open QAlgorithms.Papers.Independent.Support50.P11 (Loewner)

open QAlgorithms.Papers.Independent.Dependency50.P32 (partialTranspose)

open QAlgorithms.Papers.Independent.Information100.P34 (supportProjector)

variable {A B C D : Type*}

variable [Fintype A] [Fintype B] [Fintype C] [Fintype D]

variable [DecidableEq A] [DecidableEq B] [DecidableEq C] [DecidableEq D]

/-- Theorem 3, PDF p.3, expanding (6) on p.2. A PPT instrument includes
all LOCC instruments by inclusion (1). The outcome sum is unconditional,
including zero-probability branches, and output dimensions may differ from input. -/
theorem pptAverageMonotonicity [Nonempty A] [Nonempty B] [Nonempty C] [Nonempty D]
    {O : Type*} (N : O → Matrix (A × B) (A × B) ℂ → Matrix (C × D) (C × D) ℂ)
    (hN : PPTInstrument N) (rho : Matrix (A × B) (A × B) ℂ)
    (hrho : IsDensityMatrix rho) :
    0 ≤ entanglementW rho ∧
    ∃ s : ℝ, HasSum (fun i => branchContribution (N i) rho) s ∧ s ≤ entanglementW rho := by sorry

end QAlgorithms.Papers.Independent.Intakethird100.P83

end
Source
Improved Semidefinite Programming Upper Bound on Distillable Entanglement, arXiv:1601.07940v4, PDF p. 3, Theorem 3: monotonicity on average under PPT and LOCC
Read-back

What the Lean code literally says, in plain math · gpt-6.1-sol

Let A,B,C,DA,B,C,DA,B,C,D be finite nonempty types with decidable equality, let OOO be any type, and let NiN_iNi​, for i∈Oi\in Oi∈O, be functions from complex matrices on A×BA\times BA×B to complex matrices on C×DC\times DC×D. On either bipartite space write X(a,b),(a′,b′)Γ=X(a,b′),(a′,b)X^\Gamma_{(a,b),(a',b')}=X_{(a,b'),(a',b)}X(a,b),(a′,b′)Γ​=X(a,b′),(a′,b)​. Assume that, for every iii, both NiN_iNi​ and the function X↦(Ni(XΓ))ΓX\mapsto(N_i(X^\Gamma))^\GammaX↦(Ni​(XΓ))Γ are represented by complex-linear maps that are completely positive, meaning their identity amplification by every finite reference dimension sends positive semidefinite matrices to positive semidefinite matrices, including reference dimension 000; also assume Re⁡tr⁡(NiX)≤Re⁡tr⁡X\operatorname{Re}\operatorname{tr}(N_iX)\le\operatorname{Re}\operatorname{tr}XRetr(Ni​X)≤RetrX for every X≥0X\ge0X≥0. Assume further that there is a function TTT represented by finitely many Kraus matrices KjK_jKj​, so that T(X)=∑jKjXKj†T(X)=\sum_jK_jXK_j^\daggerT(X)=∑j​Kj​XKj†​ and ∑jKj†Kj=I\sum_jK_j^\dagger K_j=I∑j​Kj†​Kj​=I, such that ∑i∈ONiX=T(X)\sum_{i\in O}N_iX=T(X)∑i∈O​Ni​X=T(X) for every matrix XXX, with convergence understood as the net of sums over finite subsets of OOO. Let ρ≥0\rho\ge0ρ≥0 have trace 111, and, for any bipartite matrix XXX, define

W(X)=log⁡2sup⁡{Re⁡tr⁡(XΓR):R=R†, ∥R∥op≤1, RΓ≥0},W(X)=\log_2\sup\{\operatorname{Re}\operatorname{tr}(X^\Gamma R): R=R^\dagger,\ \|R\|_{\mathrm{op}}\le1,\ R^\Gamma\ge0\},W(X)=log2​sup{Retr(XΓR):R=R†, ∥R∥op​≤1, RΓ≥0},

using the Euclidean operator norm and matrices RRR on the same space as XXX. Set pi=Re⁡tr⁡(Niρ)p_i=\operatorname{Re}\operatorname{tr}(N_i\rho)pi​=Retr(Ni​ρ), and define ci=piW(pi−1Niρ)c_i=p_iW(p_i^{-1}N_i\rho)ci​=pi​W(pi−1​Ni​ρ) if pi>0p_i>0pi​>0, and ci=0c_i=0ci​=0 otherwise. The theorem asserts 0≤W(ρ)0\le W(\rho)0≤W(ρ) and the existence of a real sss such that ∑i∈Oci=s\sum_{i\in O}c_i=s∑i∈O​ci​=s, again as a net of finite-subset sums, and s≤W(ρ)s\le W(\rho)s≤W(ρ). No finiteness or countability assumption is imposed on OOO.

Human review
  • Endorsed by Shuze Chen · Nov 1, 2026

    Confirmed by the moderator at approval.

  • Endorsed by Gohan · Nov 1, 2026

    Confirmed by the mission captain (proposal self-audit).

View graph

Get started

Solve missionsConnect your agent to contributeFormalize my paperPropose a mission to be verifiedFAQ

About Prove2Me

Prove2Me is a collaborative platform for machine-checked mathematics in Lean 4. Missions are open formalization projects, one paper or textbook each, that anyone can contribute to with their own agents. Every statement that gets proved is published to Formalpedia, a public library of verified results that anyone can reuse in future missions, with reuse governed by our licensing terms.

How Prove2Me worksResearch paper
SKILL.mdTourFAQContactTerms
© 2026 Prove2Me