has_subspace_compatible_incompatible_family
There exists a family of k binary measurements on C^d that is incompatible on C^d but whose compression to every proper subspace is jointly measurable.
Statement
def has_subspace_compatible_incompatible_family (d k : ℕ) : Prop
- Defined in
- arXiv:2610.37347 Corollary 1
- Verified
- 07 Nov 2026
- Axioms
- none listed
- Used by 5
- open_problem_693_three_dim_fewer_than_nine_if_lemmasThe three-effect qutrit construction solves OP-693 conditional on the explicitly stated source, phase, spectral, parent and projection…
- subspace_incompatible_family_of_mic_dataFor all d k : ℕ, let g : Fin k → Fin d → ℂ, write G i = Matrix.vecMulVec (g i) (star (g i)) (the rank-one positive matrix g_i g_i^*), let…
- subspace_incompatible_family_dimension_squaredFor every integer d >= 2, there exists a family of d^2 binary measurements on C^d that is incompatible on C^d and jointly measurable after…
- open_problem_693_three_dim_fewer_than_nineThere exists a natural number k strictly less than 9 such that has_subspace_compatible_incompatible_family 3 k holds, meaning there is a…
- not_has_subspace_compatible_incompatible_family_oneFor any dimension d, no single binary measurement can be incompatible on C^d, so has_subspace_compatible_incompatible_family d 1 does not…
Read back from the Lean
has_subspace_compatible_incompatible_family d k asserts that for given natural numbers $d$ and $k$, there exists a family of $d \times d$ complex matrices $(M_j)_{j \in \text{Fin } k}$ satisfying:
1. **Validity as quantum effects**: For every $j \in \text{Fin } k$, both $M_j$ and $I - M_j$ are positive semidefinite (where $1$ is the $d \times d$ identity matrix).
2. **Global incompatibility**: There does NOT exist a joint POVM $G : (\text{Fin } k \to \text{Bool}) \to \text{Matrix } (\text{Fin } d) (\text{Fin } d) \ \mathbb{C}$ such that:
- $G(s)$ is positive semidefinite for all $s \in (\text{Fin } k \to \text{Bool})$,
- $\sum_s G(s) = 1$, and
- for all $j \in \text{Fin } k$, the marginal sum $\sum_{s,\, s(j) = \text{true}} G(s) = M_j$.
3. **Subspace compatibility**: For every matrix $P \in \text{Matrix } (\text{Fin } d) (\text{Fin } d) \ \mathbb{C}$ that is Hermitian ($P^* = P$), idempotent ($P^2 = P$), and has rank strictly less than $d$ (i.e., $P$ is an orthogonal projection onto a proper subspace), there exists a family $G : (\text{Fin } k \to \text{Bool}) \to \text{Matrix } (\text{Fin } d) (\text{Fin } d) \ \mathbb{C}$ such that:
- $G(s)$ is positive semidefinite for all $s$,
- $\sum_s G(s) = P$, and
- for all $j \in \text{Fin } k$, the marginal sum $\sum_{s,\, s(j) = \text{true}} G(s) = P M_j P$.
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
/-- Existence of a family of k binary measurements on C^d that is incompatible globally on C^d but jointly measurable after compression to every proper subspace. -/ def has_subspace_compatible_incompatible_family (d k : ℕ) : Prop := open scoped ComplexOrder in ∃ M : Fin k → Matrix (Fin d) (Fin d) ℂ, (∀ j, (M j).PosSemidef ∧ (1 - M j).PosSemidef) ∧ (¬ ∃ G : (Fin k → Bool) → Matrix (Fin d) (Fin d) ℂ, (∀ s, (G s).PosSemidef) ∧ (∑ s, G s = 1) ∧ ∀ j : Fin k, (∑ s, if s j then G s else 0) = M j) ∧ ∀ P : Matrix (Fin d) (Fin d) ℂ, P.IsHermitian ∧ P * P = P ∧ P.rank < d → ∃ G : (Fin k → Bool) → Matrix (Fin d) (Fin d) ℂ, (∀ s, (G s).PosSemidef) ∧ (∑ s, G s = P) ∧ ∀ j : Fin k, (∑ s, if s j then G s else 0) = P * M j * P