certifies_noiseless_codes
Given $M : \mathbb{R} \to \mathrm{Fin}\,k \to \mathrm{Matrix}(\mathrm{Fin}\,d)(\mathrm{Fin}\,d)\mathbb{C}$ and $s_{\max} : \mathbb{R}$, the predicate asserts that: $1 < s_{\max}$; for all $s \in [0, s_{\max}]$ and all $j \in \mathrm{Fin}\,k$, $M(s, j)$ and $1 - M(s, j)$ are positive semidefinite; for all $s \in [0, s_{\max}]$, the family $M(s)$ is jointly measurable on $1$ via is_jointly_measurable_on if and only if $s \le 1$; and for every natural number $D$ with $d \le D$, natural number $m$, and Kraus operators $K : \mathrm{Fin}\,m \to \mathrm{Matrix}(\mathrm{Fin}\,d)(\mathrm{Fin}\,D)\mathbb{C}$ satisfying the trace-preserving condition $\sum_i K_i^\dagger K_i = 1$, letting $N(X) = \sum_i K_i X K_i^\dagger$ and $\Lambda(Y) = \sum_i K_i^\dagger Y K_i$, there exists an isometry $W \in \mathrm{Matrix}(\mathrm{Fin}\,D)(\mathrm{Fin}\,d)\mathbb{C}$ with $W^\dagger W = 1$ such that $N(W \rho W^\dagger) = \rho$ for all density matrices $\rho$ (positive semidefinite with trace 1) if and only if for all $s \in [0, s_{\max}]$, the image family under $\Lambda$ is jointly measurable on $1$ via is_jointly_measurable_on if and only if $s \le 1$.
Statement
def certifies_noiseless_codes {d k : ℕ} (M : ℝ → Fin k → Matrix (Fin d) (Fin d) ℂ) (s_max : ℝ) : Prop
- Defined in
- arXiv:2610.37347 Discussion and outlook, paragraph 3
- Verified
- 07 Nov 2026
- Axioms
- none listed
- Built on 1
- is_jointly_measurable_onA family of k binary measurements M on C^d is jointly measurable on a subspace with orthogonal projection P if there exists a parent POVM…
- Used by 2
- not_can_certify_noiseless_codes_oneFor any dimension d, a single binary measurement cannot certify noiseless quantum codes, so ¬ can_certify_noiseless_codes d 1 holds.
- can_certify_noiseless_codesA dimension d admits certification of noiseless quantum codes by k binary measurements if there exist a parametrized family of measurement…
Read back from the Lean
Given natural numbers d and k (implicit), a family of matrix-valued functions M : ℝ → Fin k → Matrix (Fin d) (Fin d) ℂ, and a real parameter s_max : ℝ, certifies_noiseless_codes M s_max asserts that:
1. 1 < s_max;
2. For all s ∈ [0, s_max] and all j : Fin k, the matrix M s j is positive semidefinite and 1 - M s j is positive semidefinite (where 1 is the d × d identity matrix);
3. For all s ∈ [0, s_max], the family M s is jointly measurable on the identity matrix 1 (of size d × d) if and only if s ≤ 1;
4. For every natural number D with d ≤ D, every natural number m, and every family of Kraus matrices K : Fin m → Matrix (Fin d) (Fin D) ℂ satisfying ∑ i, (K i)ᴴ * K i = 1 (the D × D identity matrix):
Defining the channel N(X) = ∑ i, K i * X * (K i)ᴴ (from D × D to d × d matrices) and its dual map Λ(Y) = ∑ i, (K i)ᴴ * Y * K i (from d × d to D × D matrices),
there exists an isometry W : Matrix (Fin D) (Fin d) ℂ such that Wᴴ * W = 1 (the d × d identity matrix) and for every density matrix ρ : Matrix (Fin d) (Fin d) ℂ (i.e., ρ is positive semidefinite and has trace 1), N (W * ρ * Wᴴ) = ρ,
if and only if
for all s ∈ [0, s_max], the family j ↦ Λ (M s j) is jointly measurable on the identity matrix 1 (of size D × D) if and only if s ≤ 1.
Written by a model that saw only the Lean, never the English above. If the two disagree, that is worth a challenge.
Lean source view module on GitHub
/-- A parametrized family of k binary measurements on C^d with parameter bound s_max > 1 certifies noiseless quantum codes if its compatibility threshold is 1 and preservation of this threshold under any trace-preserving completely positive map N with Kraus operators K is equivalent to the existence of a noiseless code isometry W. -/ def certifies_noiseless_codes {d k : ℕ} (M : ℝ → Fin k → Matrix (Fin d) (Fin d) ℂ) (s_max : ℝ) : Prop := open scoped ComplexOrder Matrix in 1 < s_max ∧ (∀ s ∈ Set.Icc 0 s_max, ∀ j : Fin k, (M s j).PosSemidef ∧ (1 - M s j).PosSemidef) ∧ (∀ s ∈ Set.Icc 0 s_max, is_jointly_measurable_on (M s) 1 ↔ s ≤ 1) ∧ ∀ (D : ℕ) (_ : d ≤ D) (m : ℕ) (K : Fin m → Matrix (Fin d) (Fin D) ℂ) (_hK : ∑ i, (K i)ᴴ * K i = 1), let N : Matrix (Fin D) (Fin D) ℂ → Matrix (Fin d) (Fin d) ℂ := fun X => ∑ i, K i * X * (K i)ᴴ let Λ : Matrix (Fin d) (Fin d) ℂ → Matrix (Fin D) (Fin D) ℂ := fun Y => ∑ i, (K i)ᴴ * Y * K i (∃ W : Matrix (Fin D) (Fin d) ℂ, Wᴴ * W = 1 ∧ ∀ ρ : Matrix (Fin d) (Fin d) ℂ, ρ.PosSemidef ∧ ρ.trace = 1 → N (W * ρ * Wᴴ) = ρ) ↔ (∀ s ∈ Set.Icc 0 s_max, is_jointly_measurable_on (fun j => Λ (M s j)) 1 ↔ s ≤ 1)