Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Maximum variance of the surprisal

Open
QAlgorithms.Papers.Independent.Support50.P15.theorem8

by Goku · Oct 23, 2026 · Mathlib 0df444a (Lean v4.33.1)

prove2quantum

For a normalized state ρ\rhoρ in dimension ddd, let V(ρ)=∑ipilog⁡2pi−S(ρ)2V(\rho)=\sum_i p_i\log^2p_i-S(\rho)^2V(ρ)=∑i​pi​log2pi​−S(ρ)2 be the variance of its surprisal, where pip_ipi​ are its eigenvalues and zero eigenvalues contribute zero. Let N(d)N(d)N(d) and NdN_dNd​ be Eqs. (7)--(8). For d≥2d\ge2d≥2,

V(ρ)≤N(d)<Nd.V(\rho)\le N(d)<N_d.V(ρ)≤N(d)<Nd​.

There is exactly one maximizing rd∈(0,1/2)r_d\in(0,1/2)rd​∈(0,1/2) in Eq. (7), and equality holds if and only if the eigenvalues, with multiplicities and in any order, are 1−rd,rd/(d−1),…,rd/(d−1)1-r_d,r_d/(d-1),\ldots,r_d/(d-1)1−rd​,rd​/(d−1),…,rd​/(d−1). In dimension one V(ρ)=0V(\rho)=0V(ρ)=0. This gives the sharp dimension-only bound and its complete spectral equality characterization.

Preamble
import Definitions.Def_QAlgorithms_Papers_Independent_Support50_P15
Formal statement
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)

/-- Theorem 8, PDF p.8; proof pp.24--25. Eigenvalues are matched by a permutation, never by
an imposed order. The one-dimensional case is retained. -/
theorem theorem8 :
    (∀ a : EntropyState 1, surprisalVariance a.val = 0) ∧
    (∀ d : ℕ, 2 ≤ d →
      (∃! r : ℝ, r ∈ Set.Ioo 0 (1 / 2) ∧ varianceObjective d r = maxVariance d) ∧
      optimalMass d ∈ Set.Ioo 0 (1 / 2) ∧ varianceObjective d (optimalMass d) = maxVariance d ∧
      (∀ a : EntropyState d, surprisalVariance a.val ≤ maxVariance d ∧ maxVariance d < easyVariance d ∧
        (surprisalVariance a.val = maxVariance d ↔ ∃ e : Equiv.Perm (Fin d),
          ∀ i, ascEigenvalues a.val i = familyProb d (optimalMass d)
            (e (Fin.cast (Fintype.card_fin d) i))))) := by sorry

end QAlgorithms.Papers.Independent.Support50.P15

end
Source
Tight bound on relative entropy by entropy difference, arXiv:1304.0036v3, PDF p. 8, theorem8
Read-back

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

For a positive semidefinite complex matrix AAA of trace 111, define

S(A)=−Re⁡Tr⁡[(u↦ulog⁡u)(A)],V(A)=Re⁡Tr⁡[(u↦u(log⁡u)2)(A)]−S(A)2,S(A)=-\operatorname{Re}\operatorname{Tr}[(u\mapsto u\log u)(A)],\qquad \mathcal V(A)=\operatorname{Re}\operatorname{Tr}[(u\mapsto u(\log u)^2)(A)]-S(A)^2,S(A)=−ReTr[(u↦ulogu)(A)],V(A)=ReTr[(u↦u(logu)2)(A)]−S(A)2,

using spectral functional calculus and 0log⁡0=00\log0=00log0=0. Every 1×11\times11×1 density matrix AAA satisfies V(A)=0\mathcal V(A)=0V(A)=0. For every natural d≥2d\ge2d≥2, define

vd(r)=r(1−r)[log⁡ ⁣(1−rr(d−1))]2,Vd=sup⁡0<r<1/2vd(r),Ed=[log⁡(d−1)]24+1,v_d(r)=r(1-r)\left[\log\!\left(\frac{1-r}{r}(d-1)\right)\right]^2,\qquad V_d=\sup_{0<r<1/2}v_d(r),\qquad E_d=\frac{[\log(d-1)]^2}{4}+1,vd​(r)=r(1−r)[log(r1−r​(d−1))]2,Vd​=0<r<1/2sup​vd​(r),Ed​=4[log(d−1)]2​+1,

and let rdr_drd​ be the classically chosen real satisfying 0<rd<1/20<r_d<1/20<rd​<1/2 and vd(rd)=Vdv_d(r_d)=V_dvd​(rd​)=Vd​, if such a real exists. There exists exactly one real rrr satisfying these two conditions; the chosen rdr_drd​ satisfies them. For every d×dd\times dd×d density matrix AAA, V(A)≤Vd<Ed\mathcal V(A)\le V_d<E_dV(A)≤Vd​<Ed​, and V(A)=Vd\mathcal V(A)=V_dV(A)=Vd​ if and only if there exists a permutation π\piπ of {0,…,d−1}\{0,\ldots,d-1\}{0,…,d−1} such that its ascending eigenvalue list, including multiplicities, satisfies

λi(A)={1−rd,π(i)=0,rd/(d−1),π(i)≠0for every i.\lambda_i(A)= \begin{cases} 1-r_d,&\pi(i)=0,\\ r_d/(d-1),&\pi(i)\ne0 \end{cases} \quad\text{for every }i.λi​(A)={1−rd​,rd​/(d−1),​π(i)=0,π(i)=0​for every i.
Human review
  • Endorsed by Shuze Chen · Oct 24, 2026

    Confirmed by the moderator at approval.

  • Endorsed by Goku · Oct 24, 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