Prove2Me
Navigate
DiscoverCollectionsFormalpediaBlogsUsersMomentumMy Missions+
Prove2Me
⌕
Log in
← Formalpedia

Exact and robust self-testing of entangled two-qubit states

Open
QAlgorithms.Papers.Independent.Intakethird100.P75.bell_exact_and_robust

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

prove2quantum

Let 0≤θ≤π/40\leq\theta\leq\pi/40≤θ≤π/4. Define the unit target and Bell parameter by

∣ψθ⟩=cos⁡θ∣00⟩+sin⁡θ∣11⟩,α(θ)=2cos⁡(2θ)1+sin⁡2(2θ),b(α)=8+2α2.|\psi_\theta\rangle=\cos\theta|00\rangle+\sin\theta|11\rangle,\qquad \alpha(\theta)=\frac{2\cos(2\theta)}{\sqrt{1+\sin^2(2\theta)}},\qquad b(\alpha)=\sqrt{8+2\alpha^2}.∣ψθ​⟩=cosθ∣00⟩+sinθ∣11⟩,α(θ)=1+sin2(2θ)​2cos(2θ)​,b(α)=8+2α2​.

Then 0≤α(θ)≤20\leq\alpha(\theta)\leq20≤α(θ)≤2. The concrete qubit strategy A0=Z,A1=XA_0=Z,A_1=XA0​=Z,A1​=X, B0=cos⁡μZ+sin⁡μX,B1=cos⁡μZ−sin⁡μXB_0=\cos\mu Z+\sin\mu X,B_1=\cos\mu Z-\sin\mu XB0​=cosμZ+sinμX,B1​=cosμZ−sinμX, with Pauli matrices X,ZX,ZX,Z and μ=arctan⁡(sin⁡(2θ))\mu=\arctan(\sin(2\theta))μ=arctan(sin(2θ)), consists of dichotomic observables and attains b(α(θ))b(\alpha(\theta))b(α(θ)) on this target.

Let HA,HBH_A,H_BHA​,HB​ be arbitrary complex Hilbert spaces, ∣ψ′⟩|\psi'\rangle∣ψ′⟩ a unit vector in their completed Hilbert tensor product, and A0,A1,B0,B1A_0,A_1,B_0,B_1A0​,A1​,B0​,B1​ bounded self-adjoint involutions on the indicated local spaces. Set

Bα=αA0⊗I+A0⊗(B0+B1)+A1⊗(B0−B1).\mathcal B_\alpha=\alpha A_0\otimes I+ A_0\otimes(B_0+B_1)+A_1\otimes(B_0-B_1).Bα​=αA0​⊗I+A0​⊗(B0​+B1​)+A1​⊗(B0​−B1​).

If its expectation on ψ′\psi'ψ′ equals b(α(θ))b(\alpha(\theta))b(α(θ)), there exist local complex-linear isometric embeddings ΦA,ΦB\Phi_A,\Phi_BΦA​,ΦB​ and normalized, possibly entangled junk with

(ΦA⊗ΦB)∣ψ′⟩=∣junk⟩⊗∣ψθ⟩.(\Phi_A\otimes\Phi_B)|\psi'\rangle =|\mathrm{junk}\rangle\otimes|\psi_\theta\rangle.(ΦA​⊗ΦB​)∣ψ′⟩=∣junk⟩⊗∣ψθ​⟩.

For each fixed 0<θ≤π/40<\theta\leq\pi/40<θ≤π/4, for every η>0\eta>0η>0 there is a δ>0\delta>0δ>0, depending only on θ,η\theta,\etaθ,η, such that every realization above with Bell expectation at least b(α(θ))−δb(\alpha(\theta))-\deltab(α(θ))−δ admits such embeddings and junk satisfying

∥(ΦA⊗ΦB)∣ψ′⟩−∣junk⟩⊗∣ψθ⟩∥≤η.\|(\Phi_A\otimes\Phi_B)|\psi'\rangle- |\mathrm{junk}\rangle\otimes|\psi_\theta\rangle\|\leq\eta.∥(ΦA​⊗ΦB​)∣ψ′⟩−∣junk⟩⊗∣ψθ​⟩∥≤η.

Equalities and norms use one fixed regrouping of physical and added registers. This certifies the state, without a measurement-intertwining claim. The endpoint θ=0,α=2\theta=0,\alpha=2θ=0,α=2 is retained in the exact clause as product extraction.

Formalization Note The printed tangent relation is corrected to a double-angle relation via the continuous parameter formula above. In particular, α=0\alpha=0α=0 gives θ=π/4\theta=\pi/4θ=π/4. The qualitative robustness clause does not assert the defective explicit coefficients.

Preamble
import Definitions.Def_QAlgorithms_Papers_Independent_Intakethird100_P75
Formal statement
noncomputable section

open scoped BigOperators

namespace QAlgorithms.Papers.Independent.Intakethird100.P75

universe u v w x

open QAlgorithms.Papers.Independent.Claimed50.P34 (HilbertTensor)

open QAlgorithms.Papers.Independent.Renew46.P03 (tensorVector)

/-- PDF pp.2–3, unnumbered Theorem, with robustness in the sense of (3).
The printed Schmidt relation is corrected as documented at tiltParameter.
The ideal qubit strategy realizes every target nonvacuously. Exact extraction
includes theta=0,alpha=2 as the product branch. Robustness is fixed-theta and
uniform over arbitrary physical Hilbert spaces and all dichotomic observables.
No measurement-intertwining conclusion or invalid (19)–(20) estimate is added. -/
theorem bell_exact_and_robust (theta : ℝ) (htheta : 0 ≤ theta)
    (hthetaUpper : theta ≤ Real.pi / 4) :
    0 ≤ tiltParameter theta ∧ tiltParameter theta ≤ 2 ∧
    ‖schmidtTarget (qubitCoefficients theta)‖ = 1 ∧
    Dichotomic idealZ ∧ Dichotomic idealX ∧
    Dichotomic (idealB0 theta) ∧ Dichotomic (idealB1 theta) ∧
    bellValue (tiltParameter theta) idealZ idealX (idealB0 theta) (idealB1 theta)
      (schmidtTarget (qubitCoefficients theta)) = quantumMaximum (tiltParameter theta) ∧
    (∀ (H : Type u) (K : Type v)
      [NormedAddCommGroup H] [InnerProductSpace ℂ H] [CompleteSpace H]
      [NormedAddCommGroup K] [InnerProductSpace ℂ K] [CompleteSpace K]
      (psi : HilbertTensor H K) (A0 A1 : H →L[ℂ] H) (B0 B1 : K →L[ℂ] K),
      ‖psi‖ = 1 → Dichotomic A0 → Dichotomic A1 → Dichotomic B0 → Dichotomic B1 →
      bellValue (tiltParameter theta) A0 A1 B0 B1 psi =
        quantumMaximum (tiltParameter theta) →
      ExtractsWithin (qubitCoefficients theta) psi 0) ∧
    (0 < theta → ∀ eta : ℝ, 0 < eta → ∃ delta : ℝ, 0 < delta ∧
      ∀ (H : Type u) (K : Type v)
      [NormedAddCommGroup H] [InnerProductSpace ℂ H] [CompleteSpace H]
      [NormedAddCommGroup K] [InnerProductSpace ℂ K] [CompleteSpace K]
      (psi : HilbertTensor H K) (A0 A1 : H →L[ℂ] H) (B0 B1 : K →L[ℂ] K),
      ‖psi‖ = 1 → Dichotomic A0 → Dichotomic A1 → Dichotomic B0 → Dichotomic B1 →
      quantumMaximum (tiltParameter theta) - delta ≤
        bellValue (tiltParameter theta) A0 A1 B0 B1 psi →
      ExtractsWithin (qubitCoefficients theta) psi eta) := by sorry

end QAlgorithms.Papers.Independent.Intakethird100.P75

end
Source
Robust Self Testing of Unknown Quantum Systems into Any Entangled Two-Qubit States, arXiv:1210.4409v2, PDF p. 2, §III, unnumbered Theorem: exact and robust self-testing of every pure entangled two-qubit state
Read-back

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

For every real θ\thetaθ satisfying 0≤θ≤π/40\le\theta\le\pi/40≤θ≤π/4, put α=2cos⁡(2θ)/1+sin⁡2(2θ)\alpha=2\cos(2\theta)/\sqrt{1+\sin^2(2\theta)}α=2cos(2θ)/1+sin2(2θ)​, μ=arctan⁡(sin⁡(2θ))\mu=\arctan(\sin(2\theta))μ=arctan(sin(2θ)), Z=diag⁡(1,−1)Z=\operatorname{diag}(1,-1)Z=diag(1,−1), X=(0110)X=\begin{pmatrix}0&1\\1&0\end{pmatrix}X=(01​10​), B0=cos⁡μ Z+sin⁡μ XB_0=\cos\mu\,Z+\sin\mu\,XB0​=cosμZ+sinμX, B1=cos⁡μ Z−sin⁡μ XB_1=\cos\mu\,Z-\sin\mu\,XB1​=cosμZ−sinμX, and ϕθ=cos⁡θ e0⊗e0+sin⁡θ e1⊗e1\phi_\theta=\cos\theta\,e_0\otimes e_0+\sin\theta\,e_1\otimes e_1ϕθ​=cosθe0​⊗e0​+sinθe1​⊗e1​, where e0,e1e_0,e_1e0​,e1​ are the standard vectors indexed by false,true\mathrm{false},\mathrm{true}false,true. All tensor products here are completed complex Hilbert tensor products, namely completions of the algebraic tensor products, and tensor products of operators denote the continuous extensions of their actions on elementary tensors. The statement asserts 0≤α≤20\le\alpha\le20≤α≤2, ∥ϕθ∥=1\|\phi_\theta\|=1∥ϕθ​∥=1, and that each of Z,X,B0,B1Z,X,B_0,B_1Z,X,B0​,B1​ is self-adjoint and has square equal to the identity. Define

Bα(A0,A1,C0,C1)=αA0⊗I+A0⊗(C0+C1)+A1⊗(C0−C1).\mathcal B_\alpha(A_0,A_1,C_0,C_1) =\alpha A_0\otimes I+A_0\otimes(C_0+C_1)+A_1\otimes(C_0-C_1).Bα​(A0​,A1​,C0​,C1​)=αA0​⊗I+A0​⊗(C0​+C1​)+A1​⊗(C0​−C1​).

Then Re⁡⟨ϕθ,Bα(Z,X,B0,B1)ϕθ⟩=8+2α2\operatorname{Re}\langle\phi_\theta,\mathcal B_\alpha(Z,X,B_0,B_1)\phi_\theta\rangle=\sqrt{8+2\alpha^2}Re⟨ϕθ​,Bα​(Z,X,B0​,B1​)ϕθ​⟩=8+2α2​. Moreover, for every pair of complete complex normed inner product spaces H,KH,KH,K, every ψ∈H⊗^K\psi\in H\widehat\otimes Kψ∈H⊗​K with ∥ψ∥=1\|\psi\|=1∥ψ∥=1, and every continuous complex-linear A0,A1:H→HA_0,A_1:H\to HA0​,A1​:H→H and C0,C1:K→KC_0,C_1:K\to KC0​,C1​:K→K that are self-adjoint and square to the identity, equality Re⁡⟨ψ,Bα(A0,A1,C0,C1)ψ⟩=8+2α2\operatorname{Re}\langle\psi,\mathcal B_\alpha(A_0,A_1,C_0,C_1)\psi\rangle=\sqrt{8+2\alpha^2}Re⟨ψ,Bα​(A0​,A1​,C0​,C1​)ψ⟩=8+2α2​ implies that there exist continuous complex-linear isometries FA:H→H⊗^C2F_A:H\to H\widehat\otimes\mathbb C^2FA​:H→H⊗​C2 and FB:K→K⊗^C2F_B:K\to K\widehat\otimes\mathbb C^2FB​:K→K⊗​C2 and a vector J∈H⊗^KJ\in H\widehat\otimes KJ∈H⊗​K with ∥J∥=1\|J\|=1∥J∥=1 such that (FA⊗FB)ψ=cos⁡θ T0J+sin⁡θ T1J(F_A\otimes F_B)\psi=\cos\theta\,T_0J+\sin\theta\,T_1J(FA​⊗FB​)ψ=cosθT0​J+sinθT1​J, where TbT_bTb​ is the continuous extension of h⊗k↦(h⊗eb)⊗(k⊗eb)h\otimes k\mapsto(h\otimes e_b)\otimes(k\otimes e_b)h⊗k↦(h⊗eb​)⊗(k⊗eb​). Finally, if θ>0\theta>0θ>0, then for every real η>0\eta>0η>0 there exists a real δ>0\delta>0δ>0 such that, uniformly for all such H,K,ψ,A0,A1,C0,C1H,K,\psi,A_0,A_1,C_0,C_1H,K,ψ,A0​,A1​,C0​,C1​, the inequality 8+2α2−δ≤Re⁡⟨ψ,Bα(A0,A1,C0,C1)ψ⟩\sqrt{8+2\alpha^2}-\delta\le\operatorname{Re}\langle\psi,\mathcal B_\alpha(A_0,A_1,C_0,C_1)\psi\rangle8+2α2​−δ≤Re⟨ψ,Bα​(A0​,A1​,C0​,C1​)ψ⟩ implies the existence of such isometries and unit vector JJJ satisfying ∥(FA⊗FB)ψ−cos⁡θ T0J−sin⁡θ T1J∥≤η\|(F_A\otimes F_B)\psi-\cos\theta\,T_0J-\sin\theta\,T_1J\|\le\eta∥(FA​⊗FB​)ψ−cosθT0​J−sinθT1​J∥≤η. The exact assertion includes θ=0\theta=0θ=0; the final robustness assertion is conditional on θ>0\theta>0θ>0.

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