Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

§§3.3–3.3.1: root-fidelity curvature and extension

Open
QAlgorithms.Papers.Independent.Intakenext100.P07.root_fidelity_properties

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)∣.

Root algebraic fidelity is jointly concave on positive operators. Squared algebraic fidelity is nondecreasing under CPTP maps and obeys the displayed congruence identity. For subnormalized states, adjoining a vacuum flag of mass 1-Tr rho defines the normalized hat state; generalized fidelity has the displayed hat identity. Its square root is jointly concave, and generalized fidelity is separately concave in each argument.

F is jointly concave,F(LρL†,τ)=F(ρ,L†τL),F∗(ρ,τ)=F(ρ^,τ^).\sqrt{F}\text{ is jointly concave},\quad F(L\rho L^\dagger,\tau)=F(\rho,L^\dagger\tau L),\quad F_*(\rho,\tau)=F(\widehat\rho,\widehat\tau).F​ is jointly concave,F(LρL†,τ)=F(ρ,L†τL),F∗​(ρ,τ)=F(ρ​,τ).

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#3.fidelityproperties
§§3.3–3.3.1: root-fidelity curvature and extension, arXiv:1504.00233v5, PDF p.44,45,46. -/
theorem root_fidelity_properties [Nonempty I] [Nonempty J] :
    (∀ r s t u : paperMat I, r.PosSemidef → s.PosSemidef → t.PosSemidef → u.PosSemidef →
      ∀ l : ℝ, l ∈ Set.Icc 0 1 → l*rootFidelity r s+(1-l)*rootFidelity t u ≤ rootFidelity (paperMix l r t) (paperMix l s u)) ∧
    (∀ E : paperMat I →ₗ[ℂ] paperMat J, paperCPTP E → ∀ r s : paperMat I, r.PosSemidef → s.PosSemidef → paperFidelity r s ≤ paperFidelity (E r) (E s)) ∧
    (∀ r s L : paperMat I, r.PosSemidef → s.PosSemidef → paperFidelity (L*r*Lᴴ) s = paperFidelity r (Lᴴ*s*L)) ∧
    (∀ r s : paperMat I, paperSubnormalized r → paperSubnormalized s →
      paperGeneralizedFidelity r s = paperGeneralizedFidelity (hatState r) (hatState s) ∧ paperGeneralizedFidelity r s = paperFidelity (hatState r) (hatState s)) ∧
    (∀ r s t u : paperMat I, paperSubnormalized r → paperSubnormalized s → paperSubnormalized t → paperSubnormalized u →
      ∀ l : ℝ, l ∈ Set.Icc 0 1 →
      l*Real.sqrt (paperGeneralizedFidelity r s)+(1-l)*Real.sqrt (paperGeneralizedFidelity t u) ≤ Real.sqrt (paperGeneralizedFidelity (paperMix l r t) (paperMix l s u))) ∧
    (∀ r s t : paperMat I, paperSubnormalized r → paperSubnormalized s → paperSubnormalized t → ∀ l : ℝ, l ∈ Set.Icc 0 1 →
      l*paperGeneralizedFidelity r t+(1-l)*paperGeneralizedFidelity s t ≤ paperGeneralizedFidelity (paperMix l r s) t ∧
      l*paperGeneralizedFidelity t r+(1-l)*paperGeneralizedFidelity t s ≤ paperGeneralizedFidelity t (paperMix l r s)) := by sorry

end QAlgorithms.Papers.Independent.Intakenext100.P07

end
Source
Quantum Information Processing with Finite Resources -- Mathematical Foundations, arXiv:1504.00233v5, PDF p. 44, §§3.3–3.3.1: root-fidelity curvature and extension
Read-back

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

For finite nonempty index types I,JI,JI,J, with decidable equality, define

f(r,s)=Re⁡tr⁡r sr,F(r,s)=f(r,s)2,F‾(r,s)=(f(r,s)+(1−Re⁡tr⁡r)(1−Re⁡tr⁡s))2,f(r,s)=\operatorname{Re}\operatorname{tr}\sqrt{\sqrt r\,s\sqrt r}, \qquad F(r,s)=f(r,s)^2, \qquad \overline F(r,s)= \left(f(r,s)+\sqrt{(1-\operatorname{Re}\operatorname{tr}r)(1-\operatorname{Re}\operatorname{tr}s)}\right)^2,f(r,s)=Retrr​sr​​,F(r,s)=f(r,s)2,F(r,s)=(f(r,s)+(1−Retrr)(1−Retrs)​)2,

using spectral functional calculus for matrix square roots. Call rrr subnormalized when it is positive semidefinite and Re⁡tr⁡r≤1\operatorname{Re}\operatorname{tr}r\le1Retrr≤1, and write r^=r⊕[1−Re⁡tr⁡r]\widehat r=r\oplus[1-\operatorname{Re}\operatorname{tr}r]r=r⊕[1−Retrr], with zero off-diagonal blocks and one additional coordinate. Then all of the following hold: for every positive semidefinite r,s,t,ur,s,t,ur,s,t,u on III and every l∈[0,1]l\in[0,1]l∈[0,1], lf(r,s)+(1−l)f(t,u)≤f(lr+(1−l)t,ls+(1−l)u)l f(r,s)+(1-l)f(t,u)\le f(lr+(1-l)t,ls+(1-l)u)lf(r,s)+(1−l)f(t,u)≤f(lr+(1−l)t,ls+(1−l)u); for every complex-linear map EEE from matrices on III to matrices on JJJ that is completely positive and satisfies tr⁡(E(M))=tr⁡M\operatorname{tr}(E(M))=\operatorname{tr}Mtr(E(M))=trM for every matrix MMM, and every positive semidefinite r,sr,sr,s, F(r,s)≤F(E(r),E(s))F(r,s)\le F(E(r),E(s))F(r,s)≤F(E(r),E(s)); for every positive semidefinite r,sr,sr,s and arbitrary matrix LLL on III, F(LrL†,s)=F(r,L†sL)F(LrL^\dagger,s)=F(r,L^\dagger sL)F(LrL†,s)=F(r,L†sL); for every subnormalized r,sr,sr,s, both F‾(r,s)=F‾(r^,s^)\overline F(r,s)=\overline F(\widehat r,\widehat s)F(r,s)=F(r,s) and F‾(r,s)=F(r^,s^)\overline F(r,s)=F(\widehat r,\widehat s)F(r,s)=F(r,s); for every subnormalized r,s,t,ur,s,t,ur,s,t,u and every l∈[0,1]l\in[0,1]l∈[0,1],

lF‾(r,s)+(1−l)F‾(t,u)≤F‾(lr+(1−l)t,ls+(1−l)u);l\sqrt{\overline F(r,s)}+(1-l)\sqrt{\overline F(t,u)} \le \sqrt{\overline F(lr+(1-l)t,ls+(1-l)u)};lF(r,s)​+(1−l)F(t,u)​≤F(lr+(1−l)t,ls+(1−l)u)​;

and for every subnormalized r,s,tr,s,tr,s,t and every l∈[0,1]l\in[0,1]l∈[0,1], both

lF‾(r,t)+(1−l)F‾(s,t)≤F‾(lr+(1−l)s,t),lF‾(t,r)+(1−l)F‾(t,s)≤F‾(t,lr+(1−l)s).l\overline F(r,t)+(1-l)\overline F(s,t) \le\overline F(lr+(1-l)s,t), \qquad l\overline F(t,r)+(1-l)\overline F(t,s) \le\overline F(t,lr+(1-l)s).lF(r,t)+(1−l)F(s,t)≤F(lr+(1−l)s,t),lF(t,r)+(1−l)F(t,s)≤F(t,lr+(1−l)s).

Complete positivity here means that applying EEE to each block of any positive semidefinite matrix with any natural number of block indices, including zero, preserves positive semidefiniteness. Zero matrices are allowed among the positive semidefinite and subnormalized inputs, and every interval condition includes l=0l=0l=0 and l=1l=1l=1.

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