Lemma 7.3 — local spectral filtering
OpenNuclearTensor.Membership.Supplement.lemma_7_3computational-complexityp2o-batch-stoc27sp2o-gran-per-chapterp2o-plan-paperp2o-v1quantum-informationseparability
For an unknown -party state of local dimension , let , , and . The ramp is zero up to , equals between and , and is one from onward. For every , a routine implements
on a specified target register with trace-norm channel error at most , even with an arbitrary entangled reference. It uses fresh auxiliary copies and time. This supports the dimensionality-reduction companion theorem.
Formalization Note The flag is an extra orthogonal dimension; the successful and flagged blocks together are trace preserving. The unknown filter is defined spectrally but is not an executable primitive. The approximating program accesses only by fresh copies.
Differences from the source:
- encoding: finite complex Hilbert spaces and Hilbert–Schmidt geometry → row-major complex matrices on Fin (d^k), explicit pure-product hull, and an extended-nonnegative Frobenius-distance infimum; reason: these coordinates cover every finite-dimensional source instance; no density or norm restriction is added to the classical Hermitian input
- reading: algorithms with classical arithmetic and quantum copy access → finite arithmetic/spectral programs and elementary H,T,CX programs with Born-rule execution, charged fresh copies, and explicit classical outputs; reason: makes the computational model and all resource claims operational; neither global optimization nor a desired quantum subroutine is a primitive
- encoding: O and polynomial resource notation → positive integer multiplicative constants and degrees quantified before varying instance parameters; reason: preserves the indicated parameter dependencies and ties each bound to actual executed charges
- reading: cutoff f_Theta(k^2/eta^4) → cutoff a=2048k^2/eta^4 from the proof on page 31; reason: fixes the source proof parameter without changing either promise or any resource dependence
- encoding: logarithmic dimension factors → ell(d)=1+ceil(log_2 d); reason: the usual positive input-size convention includes one-dimensional systems without a zero runtime bound
Preamble
import Definitions.Def_NuclearTensor_Membership_Supplement_Tasks set_option autoImplicit false noncomputable section
Formal statement
namespace NuclearTensor.Membership.Supplement
theorem lemma_7_3 :
∃ P : QProgram, ∀ (σ : Spectral) (A r : ℕ), 0 < A → 0 < r →
∃ B C : ℕ, 0 < B ∧ 0 < C ∧
∀ d k : ℕ, 0 < d → 0 < k → ∀ η δ : ℝ, 0 < η → 0 < δ →
∀ j : Fin k, ∀ ρ : CMat (d ^ k), IsDensity ρ →
implements ρ σ A r P [d, k, j.val] [η, δ]
(flagged (localFilter ρ η j)) δ
(B * ((k : ℝ) + inverseScale η + inverseScale δ) ^ C)
(B * ((k : ℝ) + inverseScale η + inverseScale δ + logSize d) ^ C) := by sorry
end NuclearTensor.Membership.Supplement
Source
Martino Bernasconi, Giulio Malavolta, STOC 2027 submission #945 (October 2026), p. 30, Lemma 7.3; supporting proof in pp. 46–48.