Lemma 1 (Invariance from equivariance)
ProvedQAlgorithms.Papers.Independent.Support50.P30.loss_invarianceLet be a compact topological group with a continuous unitary representation on the -qubit space. Let , where every is Hermitian and commutes with every . Suppose that the Hermitian observable also commutes with every . For every parameter vector, density matrix , and ,
Thus composing the equivariant circuit with the equivariant measurement gives an invariant loss. No assumption about learning labels is required.
import Definitions.Def_QAlgorithms_Papers_Independent_Support50_P30
noncomputable section
open MeasureTheory
open scoped BigOperators ComplexConjugate Matrix.Norms.L2Operator NNReal ENNReal
namespace QAlgorithms.Papers.Independent.Support50.P30
/-- PDF p.3, Lemma 1; proof on p.22. -/
theorem loss_invariance {n L : ℕ} {G : Type*} [Group G] [TopologicalSpace G]
[IsTopologicalGroup G] [CompactSpace G]
(R : G →* Matrix.unitaryGroup (Qubits n) ℂ) (hR : Continuous R)
(H : Fin L → QnnOperator n) (hH : ∀ j, (H j).IsHermitian)
(hcomm : ∀ j g, H j * (R g).val = (R g).val * H j)
(O : QnnOperator n) (hO : O.IsHermitian)
(hOcomm : ∀ g, O * (R g).val = (R g).val * O) :
∀ (theta : Fin L → ℝ) (rho : QnnDensity n) (g : G),
qnnLoss (circuit H theta) ((R g).val * rho.val * (R g).val.conjTranspose) O =
qnnLoss (circuit H theta) rho.val O := by sorry
end QAlgorithms.Papers.Independent.Support50.P30
end
Read-back
What the Lean code literally says, in plain math · gpt-6.1-sol
For every pair of natural numbers , including zero, let . Let be a group equipped with a topology making it a compact topological group, and let be a continuous group homomorphism. Suppose are Hermitian matrices satisfying for every , and is a Hermitian matrix satisfying for every . For any real parameter vector , define
with the empty product equal to the identity when . Then for every positive semidefinite matrix on with , and every ,
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.