SigmaBarIndex #
Ported from scottnarmstrong/Superdiffusion at 7f20476385d4af418f110303085ad0740c9eba46. Module imports, support namespace and local option directives adapted for Lean Pool.
The σ index conversion at a two-scale gap, binder-free #
This module closes that conversion inside the anchor's own regime: the
induction-state binder that Annular.SigmaBarBudget.sigmaBar_ratio_le_four
carries is discharged internally by
GoodEvents.exists_allScalesInductionState, at the landmark m₀ := n+2 (resp.
m₀ := n), so nothing beyond γ ≤ C^{-1} c⋆^{10} is asked of the caller.
This is the B1 wrapper pattern of Provider/BoundsEaL/StepFourSigmaBar.lean,
re-derived here so that the BoundsEaL tree is not imported.
σ_{n+2}^{-1} ≤ 4 σ_n^{-1} (the direction the anchor's display needs)
σ_n^{-1} ≤ 7 σ_{n+2}^{-1} (the converse, at the same regime)
Both constants are absolute numerals: no s, no γ, no c⋆, no dimension.
The regime is written in the frozen theorem's own shape γ ≤ C^{-1} c⋆^{10}
(the induction-state producer's shape γ ≤ (C₀^{-1})^{10} c⋆^{10} is reached
by exporting C := C₀^{10}).
Where the constants come from #
σ is almost monotone in the scale, in both directions, because the
induction state pins σ_m^2 inside a factor-4 window around max(c⋆ γ^{-1} 3^{2γm}, ν^2), and that window's own ratio across two scales is 3^{4γ} ≤ 3
at the standing γ ≤ 1/4. Upward: the window is monotone in m, so the two
factor-4 branches give σ_n ≤ 4 σ_{n+2}. Downward: the extra 3^{4γ} ≤ 3
gives σ_{n+2}^2 ≤ 48 σ_n^2, i.e. σ_{n+2} ≤ 7 σ_n.
References #
- ABK26,
e.shom.h.bounds;p.induction.boundsproof.
1. Two arithmetic helpers #
2. The two ratio readings of the state #
σ is almost increasing. From the induction state at the landmark
n+2: σ_n ≤ 4 σ_{n+2}, in the inverse form the force leg consumes.
3. The binder-free exports #
The σ index conversion the anchor's force leg needs, binder-free.
In the frozen theorem's own regime alone, σ_{n+2}^{-1} ≤ 4 σ_n^{-1} at
every integer scale: the induction-state binder is discharged internally.