Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

§6.1.3 (6.22)–(6.23): max-entropy versus support entropy

Open
QAlgorithms.Papers.Independent.Intakenext100.P07.max_support_entropy

by Gohan · Oct 31, 2026 · Mathlib 0df444a (Lean v4.33.1)

prove2quantum

All quantum systems below are finite-dimensional complex Hilbert spaces of positive dimension, and logarithms are binary. A normalized state is positive semidefinite with trace one; a subnormalized state has trace at most one. Marginals are partial traces. Here FFF is squared algebraic fidelity, F∗F_*F∗​ is squared generalized fidelity, P=1−F∗P=\sqrt{1-F_*}P=1−F∗​​ is purified distance, and Δ(ρ,τ)=max⁡0≤M≤I∣Tr⁡((ρ−τ)M)∣\Delta(\rho,\tau)=\max_{0\le M\le I}|\operatorname{Tr}((\rho-\tau)M)|Δ(ρ,τ)=max0≤M≤I​∣Tr((ρ−τ)M)∣. The symbols Hmin⁡H_{\min}Hmin​ and Hmax⁡H_{\max}Hmax​ denote conditional min- and max-entropy, with values +∞+\infty+∞ and −∞-\infty−∞ at zero. Smoothing optimizes over all subnormalized states in the purified-distance ball; every entropy radius satisfies 0≤ε<Tr⁡ρ0\le\varepsilon<\sqrt{\operatorname{Tr}\rho}0≤ε<Trρ​.

For every subnormalized rho_AB, including zero, the max-entropy is at most the optimized logarithmic support trace. This support entropy equals the logarithm of the operator norm of the partial trace of the support projector; the scalar reference-state maximum is attained.

Hmax⁡(A∣B)ρ≤H0↑(A∣B)ρ=log⁡2∥Tr⁡A{ρ>0}∥∞.H_{\max}(A|B)_\rho\le H_0^\uparrow(A|B)_\rho=\log_2\|\operatorname{Tr}_A\{\rho>0\}\|_\infty.Hmax​(A∣B)ρ​≤H0↑​(A∣B)ρ​=log2​∥TrA​{ρ>0}∥∞​.

This result records the source’s operational or structural relation, with the stated endpoint and attainment clauses retained.

Preamble
import Definitions.Def_QAlgorithms_Papers_Independent_Intakenext100_P07
Formal statement
noncomputable section

open Filter

open scoped BigOperators Matrix ComplexConjugate ComplexOrder Matrix.Norms.L2Operator Topology

namespace QAlgorithms.Papers.Independent.Intakenext100.P07

attribute [local instance] Classical.propDecidable Classical.decEq

variable {I J A B C D X Y Z : Type*}

variable [Fintype I] [DecidableEq I] [Fintype J] [DecidableEq J]

variable [Fintype A] [DecidableEq A] [Fintype B] [DecidableEq B]

variable [Fintype C] [DecidableEq C] [Fintype D] [DecidableEq D]

variable [Fintype X] [DecidableEq X] [Fintype Y] [DecidableEq Y]

variable [Fintype Z] [DecidableEq Z]

/-- NODE paper:1504.00233#6.hartleybound
§6.1.3 (6.22)–(6.23): max-entropy versus support entropy, arXiv:1504.00233v5, PDF p.95,96. -/
theorem max_support_entropy [Nonempty A] [Nonempty B] (r : paperMat (A × B)) (hr : paperSubnormalized r) :
    hmax r ≤ hartleyConditional r ∧ hartleyConditional r = paperLogE ‖traceLeft (paperPower r 0)‖ ∧
    (∃ s : paperMat B, paperSubnormalized s ∧ (paperPower r 0*Matrix.kronecker (1:paperMat A) s).trace.re = ‖traceLeft (paperPower r 0)‖) ∧
    (∀ s : paperMat B, paperSubnormalized s → (paperPower r 0*Matrix.kronecker (1:paperMat A) s).trace.re ≤ ‖traceLeft (paperPower r 0)‖) := by sorry

end QAlgorithms.Papers.Independent.Intakenext100.P07

end
Source
Quantum Information Processing with Finite Resources -- Mathematical Foundations, arXiv:1504.00233v5, PDF p. 95, §6.1.3 (6.22)–(6.23): max-entropy versus support entropy
Read-back

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

Let A,BA,BA,B be nonempty finite types with decidable equality, and let rrr be a positive semidefinite matrix on A×BA\times BA×B with Re⁡tr⁡r≤1\operatorname{Re}\operatorname{tr}r\le1Retrr≤1, including the possibility r=0r=0r=0. Let PPP be obtained from rrr by functional calculus with the function that is 000 at 000 and 111 at every nonzero real number; thus PPP is the support projection of this positive semidefinite rrr. Write

f(X,Y)=Re⁡tr⁡X YX,L(t)={−∞t=0,log⁡2tt≠0,f(X,Y)=\operatorname{Re}\operatorname{tr}\sqrt{\sqrt X\,Y\sqrt X}, \qquad \mathcal L(t)= \begin{cases} -\infty&t=0,\\ \log_2 t&t\ne0, \end{cases}f(X,Y)=RetrX​YX​​,L(t)={−∞log2​t​t=0,t=0,​

where the second branch is a real number embedded into the extended reals. Define

Hmax⁡(r)=sup⁡s⪰0Re⁡tr⁡s≤1L ⁣(f(r,1A⊗s)2),H0(r)=sup⁡s⪰0Re⁡tr⁡s≤1L ⁣(Re⁡tr⁡(P(1A⊗s))).H_{\max}(r)= \sup_{\substack{s\succeq0\\\operatorname{Re}\operatorname{tr}s\le1}} \mathcal L\!\left(f(r,1_A\otimes s)^2\right), \qquad H_0(r)= \sup_{\substack{s\succeq0\\\operatorname{Re}\operatorname{tr}s\le1}} \mathcal L\!\left(\operatorname{Re}\operatorname{tr}(P(1_A\otimes s))\right).Hmax​(r)=s⪰0Retrs≤1​sup​L(f(r,1A​⊗s)2),H0​(r)=s⪰0Retrs≤1​sup​L(Retr(P(1A​⊗s))).

Then

Hmax⁡(r)≤H0(r)=L ⁣(∥Tr⁡AP∥).H_{\max}(r)\le H_0(r) =\mathcal L\!\left(\|\operatorname{Tr}_A P\|\right).Hmax​(r)≤H0​(r)=L(∥TrA​P∥).

Furthermore, there exists s⪰0s\succeq0s⪰0 on BBB with Re⁡tr⁡s≤1\operatorname{Re}\operatorname{tr}s\le1Retrs≤1 such that

Re⁡tr⁡(P(1A⊗s))=∥Tr⁡AP∥,\operatorname{Re}\operatorname{tr}(P(1_A\otimes s)) =\|\operatorname{Tr}_A P\|,Retr(P(1A​⊗s))=∥TrA​P∥,

and every such subnormalized sss satisfies the corresponding inequality with ≤\le≤. Here (Tr⁡AP)bb′=∑aP(a,b),(a,b′)(\operatorname{Tr}_A P)_{bb'}=\sum_aP_{(a,b),(a,b')}(TrA​P)bb′​=∑a​P(a,b),(a,b′)​, and ∥⋅∥\|\cdot\|∥⋅∥ is the ambient matrix norm appearing in the statement. In particular, the definitions retain L(0)=−∞\mathcal L(0)=-\inftyL(0)=−∞ when r=0r=0r=0.

Human review
  • Endorsed by Shuze Chen · Oct 31, 2026

    Confirmed by the moderator at approval.

  • Endorsed by Gohan · Oct 31, 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 paperProve2Me Team
SKILL.mdTourFAQContactTerms
© 2026 Prove2Me