Frozen learning: finite dynamics, asymptotic prediction and norm-log rate
OpenQAlgorithms.Papers.Independent.Intakenext100.P88.frozen_learningThe circuit acts on a nonzero finite-dimensional complex Hilbert space and has the ordered form , with fixed unitary , Hermitian , and real angles. Products are written in increasing index order and the rightmost factor acts first. Feature vectors have norm one. For Hermitian observables , put .
Absorb the reference angles into the fixed gates and set , where . Write and for derivatives of the actual circuit. The frozen kernel is .
Let and be nonempty finite training and observable index sets. Put , let be the first Taylor output at initial rescaled angles , and set on training indices. Assume , , , and . Equivalently, all eigenvalues obey , the positive-step geometric-series and real-log regime used in the source. Under the frozen recurrences, for every and general data-observable index ,
On training indices . The paper's statistic satisfies . This packages the finite solution, training interpolation, and prediction on general data. The norm is the L2 operator norm, and the inverse is solely the restricted training inverse.
import Definitions.Def_QAlgorithms_Papers_Independent_Intakenext100_P88
noncomputable section
namespace QAlgorithms.Papers.Independent.Intakenext100.P88
open scoped BigOperators ComplexConjugate Topology NNReal ComplexOrder
open MeasureTheory ProbabilityTheory Filter
/-- PDF p.6 Theorem 3, SM pp.24-25 Eqs.(98)-(104): frozen finite-time solution,
asymptotic prediction and training interpolation. The inverse is only the training kernel.
Positive definiteness of I-eta K and its strict L2 contraction make the proof's positive
geometric-series/logarithm regime explicit. The norm-log approximation is its one-sided
small-learning-rate leading term; it is not the slowest-mode rate. -/
theorem frozen_learning {D T I : Type*} [Fintype T] [DecidableEq T] [Nonempty T]
[Fintype I] [DecidableEq I] [Nonempty I] {h L : ℕ}
(M : ExpectationNetwork D I h L) (embed : T ↪ D)
(delta eta : ℝ) (hdelta : 0 < delta) (heta : 0 < eta)
(v : Fin L → ℝ) (y : T × I → ℝ)
(hinv : (trainingKernel M delta embed).det ≠ 0)
(hpositive : (1 - eta • trainingKernel M delta embed).PosDef)
(hcontract : QAlgorithms.specNorm (1 - eta • trainingKernel M delta embed) < 1) :
let K := trainingKernel M delta embed
let z := linearOutput M delta v
let e : EuclideanSpace ℝ (T × I) := WithLp.toLp 2
(fun a => z (trainingIndex embed a) - y a)
let Z := prediction (frozenKernel M delta) (trainingIndex embed) eta z e
(∀ t, frozenResidual K eta e t = l2Action ((1 - eta • K) ^ t) e) ∧
Tendsto (frozenResidual K eta e) atTop (nhds 0) ∧
(∀ a, Tendsto (fun t => Z t a) atTop
(nhds (z a - ∑ b, ∑ c, frozenKernel M delta a (trainingIndex embed b) *
(K⁻¹) b c * e c))) ∧
(∀ a, Tendsto (fun t => Z t (trainingIndex embed a) - y a) atTop (nhds 0)) ∧
Tendsto (fun s : ℝ => logRate K s / s) (nhdsWithin 0 (Set.Ioi 0))
(nhds (QAlgorithms.specNorm K)) := by sorry
end QAlgorithms.Papers.Independent.Intakenext100.P88
end
Read-back
What the Lean code literally says, in plain math · gpt-6.1-sol
Let be types, with and finite and nonempty and with decidable equality, and let be natural numbers. Let consist of , unitary complex matrices , Hermitian matrices , unit vectors for , and Hermitian observables for . Set in increasing index order and . Fix an injection , positive reals , , and labels . Write , , , , , , and . Assume , that is positive definite, and that its operator norm for the Euclidean norm satisfies . Define , , , and . Then for every , in Euclidean norm, and for every ,
For every , . Finally, defining , one has ; the infinite sum uses the totalized convention assigning zero when the series is not summable. Although is a natural number, makes zero and hence cannot satisfy the nonzero-determinant hypothesis on the nonempty index set .
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.