Original finite-restored marker bound is exactly an arithmetic inequality on all original tests
ProvedConnesGreen.canonical_finite_marker_lower_iff_arithmetic_testsFor the ORIGINAL selected packet and restoration packet on the same actual-zero carrier, assume the displayed original restored denominator is strictly positive and . We prove
A closed local proof of the exact original scaled marker/covariance equivalence reduces the marker bound to M_S M_S* <= (beta inverse minus one) A_F. Strict positivity supplies self-adjointness of A_F; real scalar multiplication preserves it. Accepted ORIGINAL dense-test covariance order reduces the operator inequality to supported-test quadratic inequalities, with no density assertion about neutral kernels. The real scalar/inner-product identity and the closed exact finite-restored arithmetic quadratic proof give the displayed right side. The public marker expression unfolds EXACTLY canonicalFiniteRestoredMarker, and E(g) unfolds EXACTLY physicalTestEnergy; native readback checks both against the unchanged original theorem. The selected S is never replaced by F, and no uniform endpoint neighborhood or jump-budget estimate is assumed.
import Definitions.Def_ConnesGreen_RG0_original_actors import Definitions.Def_WeilMarker_regularized_cost open Complex MeasureTheory ConnesRZ ConnesRZFrontier ConnesGreen WeilDefect WeilDefect.ConnesNative WeilDefect.MarkerStability WeilDefect.WDT13 open scoped BigOperators InnerProductSpace lp ENNReal Classical ComplexOrder Topology set_option autoImplicit false set_option maxHeartbeats 3000000 set_option backward.isDefEq.respectTransparency false noncomputable section
theorem ConnesGreen.canonical_finite_marker_lower_iff_arithmetic_tests (t : ℝ) (ht : 0 < t)
(S F : Finset CriticalZeros) (ε β : ℝ) (hβ : 0 < β) (hβ1 : β < 1)
(hA : IsStrictlyPositive (canonicalPositiveCovariance t ht + ε • 1 -
canonicalTailCovariance t ht F)) :
β • (1 : ℓ²({ρ : CriticalZeros // ρ ∈ S}, ℂ) →L[ℂ]
ℓ²({ρ : CriticalZeros // ρ ∈ S}, ℂ)) ≤ marker (canonicalPositiveCovariance t ht + ε • 1 - canonicalTailCovariance t ht F) (canonicalSelectedSynthesis t ht S) ↔
∀ g : ℝ → ℂ, SupportedTest t g →
‖(canonicalSelectedSynthesis t ht S).adjoint (sourceEmbed t (problemOneL g))‖ ^ 2 ≤
(β⁻¹ - 1) * ((weilDistribution (conv g (starInv g))).re +
‖(canonicalSelectedSynthesis t ht F).adjoint (sourceEmbed t (problemOneL g))‖ ^ 2 +
ε * ((∫ x : ℝ, ‖iteratedDeriv 1 g x‖ ^ 2) + (1 / 4 : ℝ) * (∫ x : ℝ, ‖g x‖ ^ 2))) := by sorry