Normalized negative crossings on the alive window (dyadic blocks)
ProvedG26G.Sharp.cross_windowSetting. is a compact action set spanning , is a compact generator (), is the direction enumeration, and is a GIA-FTRL run of rounds (IsGIAFTRLRun) with cumulative loss estimate , run filtration , learning rate (gbeta) and magnitude constant (gB). The filtration used below is for and (Fx). The standing hypotheses Sharp hold: lies in the unit ball, , , is the unique minimizer of on , and the instance is sharp with modulus (H.sig): on . The corruption is bounded by : a.s. (IsCorruptedG). The projective body is and the cone body is ; , , and is the index of among the reference atoms (js).
Assume is surjective and . Let and with
Then the probability that, at some round at which the reference has been since , the normalized field has a negative crossing , satisfies
with instance constants , (H.cX1, H.cXB).
The proof splits into dyadic blocks and applies the crossing envelope to the normalized field on each block, with the projective extension (G.306) controlling the absolute envelope. This is the crossing part of the alive-window bound (G.567).
Formalization Note Rounds are 0-based.
import Mathlib
import Definitions.Def_IAFTRL_LinEnv
import Definitions.Def_IAFTRL_GIAFTRL
import Definitions.Def_IAFTRL_G26_body
import Definitions.Def_IAFTRL_G26_run
import Definitions.Def_IAFTRL_G26_window
open MeasureTheory ProbabilityTheory Set IAFTRL Matrix J11
open G26G
open G26 G26R
variable {d : ℕ} {𝒜 ℬ : Set (Fin d → ℝ)} {e : ℕ → Fin d → ℚ} {θ astar : Fin d → ℝ}
variable {Ω : Type*} [mΩ : MeasurableSpace Ω] {P : Measure Ω} [IsProbabilityMeasure P]
{T : ℕ} {R : GRun d Ω}
open G26G.Sharp
variable (H : Sharp 𝒜 ℬ θ astar) (he : Function.Surjective e)
include H heomit he in
theorem G26G.Sharp.cross_window (he : Function.Surjective e) (hR : IsGIAFTRLRun P 𝒜 ℬ e T R) {C : ℝ}
(hC : 0 ≤ C) (hcor : IsCorruptedG P 𝒜 θ C T R) {m t : ℕ} {z : ℝ} (hz0 : 0 < z)
(hm : 1 ≤ m) (hmt : m ≤ t) (htT : t < T)
(hzx : (((m : ℝ) + 1) / z) ^ 8 ≤ ((m : ℝ) + 1) / ((t : ℝ) + 1))
(hlC : H.sig⁻¹ * C ≤ (m : ℝ) + 1) (htm : ((t : ℝ) + 1) / ((m : ℝ) + 1) ≤ (m : ℝ) + 1) :
P.real {ω | ∃ s, m ≤ s ∧ s ≤ t ∧ ω ∈ JN ℬ e (js ℬ e astar) R m s ∧
tauB z s < negMax (DT ℬ θ astar) (R.cumEst s ω) (cN 𝒜 R θ H.sig⁻¹ C s ω)} ≤
2 * H.cX1 * (((m : ℝ) + 1) / ((t : ℝ) + 1)) ^ 4 + 18 * H.cXB / ((m : ℝ) + 1) ^ 4 := by sorry