Lemma 1 — sharper Wallis-ratio estimate
ProvedKhintchineTwoRegime.lemma_1boolean-functionskhintchine-inequalityp2o-gran-per-chapterp2o-plan-paperp2o-v1rademacher-sums
Let be Haagerup's function (8) and . For every ,
Equivalently, with ,
Both forms are stated. The estimate turns Haagerup's inequality (9) into the explicit coefficient bound of Lemma 2.
Differences from the source:
- encoding: (1 + 1/(2s) + 1/(8s(s + 1/2)))^{−1/2} and F(u^{−2}) → 1/√(1 + 1/(2s) + 1/(8s(s + 1/2))) and F((u²)⁻¹); reason: the base is positive for s > 0, so X^{−1/2} = 1/√X; same values
Preamble
import Mathlib import Definitions.Def_KhintchineTwoRegime_haagerupF import Definitions.Def_KhintchineTwoRegime_Cconst import Definitions.Def_KhintchineTwoRegime_rFun
Formal statement
namespace KhintchineTwoRegime
/-- Lemma 1 (A sharper Wallis-ratio estimate) (Fang–Wang, *A Two-regime Khintchine Inequality and an Improved Bound on the Degree-1 Fourier Weight for Linear Threshold Functions*, arXiv:2608.27908v2 (held version of 2026-11-02, sha256 1d8d9bf1…), eq. (10), p.6): for every `s > 0`,
`F(s) ≥ C (1 + 1/(2s) + 1/(8s(s + 1/2)))^{−1/2}`; equivalently `F(u^{−2}) ≥ C r(u)` for
`u > 0`.
**Formalization Note** `X^{−1/2}` is written `1/√X` (the base is positive for `s > 0`), and
`u^{−2}` is `(u²)⁻¹`. Both printed forms are stated. -/
theorem lemma_1 :
(∀ s : ℝ, 0 < s →
Cconst / Real.sqrt (1 + 1 / (2 * s) + 1 / (8 * s * (s + 1 / 2))) ≤ haagerupF s) ∧
(∀ u : ℝ, 0 < u → Cconst * rFun u ≤ haagerupF ((u ^ 2)⁻¹)) := by sorry
end KhintchineTwoRegime
Source
Fang–Wang, A Two-regime Khintchine Inequality and an Improved Bound on the Degree-1 Fourier Weight for Linear Threshold Functions, arXiv:2608.27908v2 (held version of 2026-11-02, sha256 1d8d9bf1…), Lemma 1, eq. (10), p.6
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.