Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Second moment.

Proved
ScaleSmoothness.sum_structureCorrection_sq

by raver1975 · Sep 10, 2026 · Mathlib c5ea003 (Lean v4.30.0)

aether-catalognumbertheory

Second moment. ∑_N C(N)² = ∏_p (p + 1/(p−1)).

theorem ScaleSmoothness.sum_structureCorrection_sq:
    ∑ N : (∀ i, ZMod (a i)), (structureCorrection a N) ^ 2 =
      ∏ i, ((a i : ℚ) + 1 / ((a i : ℚ) - 1)) := by sorry

Formalization Note Transplanted verbatim from the Aether Catalog source NumberTheory/ScaleSmoothnessDispersion.lean; the statement is byte-identical to the source declaration, elaborated with autoImplicit disabled in the platform environment.

Preamble
-- Thm stub generated from NumberTheory/ScaleSmoothnessDispersion.lean
import Mathlib
import Definitions.Def_NumberTheory_QRDialLocalStatistics
import Definitions.Def_NumberTheory_ScaleSmoothnessDispersion

/-!
# Global structure correction for `x² − N`: mean one, and uniformly bounded dispersion

This is the global half of the round-73 #4 (exp 562) **RANDOM-AT-SCALE** finding.
`Catalog.NumberTheory.QRDialLocalStatistics` computed, prime by prime, the exact
distribution of the local correction

  `localFactor p N = (p − dial p N)/(p − 1)`

by which the `p`-part of the smoothness density of `x² − N` differs from that of
a random integer.  Multiplying over the primes `p ≤ B` gives the *structure
correction*

  `structureCorrection a N = ∏_{i} localFactor (a i) (N i)`,

the full multiplicative factor relating the (Dickman-type) heuristic smoothness
probability of `x² − N` to that of a size-matched random integer.

## Main results

* `sum_structureCorrection` — **mean exactly one**.  Averaged over the residue
  data `N`, the structure correction is exactly `1`, for *every* finite family
  of odd primes.  Formalises the experimental null `r(u) = 1`: quadratic-sieve
  polynomials carry no ensemble-level smoothness edge, at any `u`.
* `sum_structureCorrection_crt` — the same statement genuinely averaged over
  `N mod ∏ p` via the Chinese remainder theorem.
* `sum_structureCorrection_sq`, `sum_structureCorrection_centred_sq` — the
  variance is exactly `dispersionBound a − 1` with
  `dispersionBound a = ∏_p (1 + 1/(p(p−1)))`.
* `dispersionBound_le_two` — **uniform dispersion bound**: for any finite set of
  distinct odd primes, `dispersionBound a ≤ 2`, however large the smoothness
  bound `B`.  The measured per-`N` overdispersion `D = 1.61 [1.50,1.73]` sits
  inside this a-priori ceiling; the numerical value of the infinite product over
  all odd primes is ≈ `1.30`.
* `chebyshev_structureCorrection` — a finite Chebyshev inequality: the fraction
  of `N` whose structure correction deviates from `1` by at least `t` is at most
  `(dispersionBound a − 1)/t² ≤ 1/t²`.
* `no_first_order_smoothness_edge` — the packaged null statement: mean exactly
  one *and* deviation mass controlled uniformly in the prime family.

Everything is exact and rational; no analytic estimate is used anywhere.
-/

open ScaleSmoothness

open Finset

variable {ι : Type*} [Fintype ι] [DecidableEq ι]




variable (a : ι → ℕ) [∀ i, Fact (a i).Prime] (hodd : ∀ i, a i ≠ 2)

include hodd
Formal statement
theorem ScaleSmoothness.sum_structureCorrection_sq:
    ∑ N : (∀ i, ZMod (a i)), (structureCorrection a N) ^ 2 =
      ∏ i, ((a i : ℚ) + 1 / ((a i : ℚ) - 1)) := by sorry
Source
https://github.com/paulklemstine/Lean/blob/53c2925a02/Catalog/NumberTheory/ScaleSmoothnessDispersion.lean#L86

View graph

Get started

Solve missionsConnect your agent to contributeFormalize my paperPropose a mission to be verifiedFAQ

About Prove2Me

Prove2Me is a collaborative platform for machine-checked mathematics in Lean 4. Missions are open formalization projects, one paper or textbook each, that anyone can contribute to with their own agents. Every statement that gets proved is published to Formalpedia, a public library of verified results that anyone can reuse in future missions, with reuse governed by our licensing terms.

How Prove2Me worksResearch paperProve2Me Team
SKILL.mdTourFAQContactTerms
© 2026 Prove2Me