claimroutineinferred_dep

Expected harmonic rank drop per level

Expected harmonic rank drop per level is at most 4θ4\theta
per-level cost summed over the depth
Claim·§3.4 Linear lifts and loss of rank·inside propFull lifts detect the leaf symbol

Statement (paper)

With Δj=Hrj−Hrj+1≥0\Delta_j=H_{r_j}-H_{r_{j+1}}\ge0, for every jj one has E[Δj∣Fj]≤3θ+Hr02r0θBj\E[\Delta_j\mid\mathcal F_j]\le3\theta+H_{r_0}2^{r_0}\theta B_j, and consequently the unconditional bound EΔj≤(3+Hr02r0ε0)θ≤4θ\E\Delta_j\le\bigl(3+H_{r_0}2^{r_0}\eps_0\bigr)\theta\le4\theta holds.

In plain English

With Δj=Hrj−Hrj+1≥0\Delta_j=H_{r_j}-H_{r_{j+1}}\ge0 the drop in harmonic rank at level jj, one has E[Δj∣Fj]≤3θ+Hr02r0θBj\E[\Delta_j\mid\mathcal F_j]\le3\theta+H_{r_0}2^{r_0}\theta B_j. Taking expectations gives EΔj≤(3+Hr02r0ε0)θ≤4θ\E\Delta_j\le(3+H_{r_0}2^{r_0}\eps_0)\theta\le4\theta.

Why it's here

This is the per-level cost estimate: the block-kernel lemma limits how much rank can be lost at a generic level, and the nongeneric indicator BjB_j accounts for the exceptional case. The fresh-isomorphism claim gives EBj≤ε0\E B_j\le\eps_0, which turns the conditional bound into the uniform 4θ4\theta per level. Summing over ss levels then shows the rank survives with good probability in the next claim.
Classified routine:
A per-level expected-cost calculation that combines the block-kernel lemma (rank loss at generic levels) with the genericity bound EBj≤ε0\E B_j\le\eps_0 from the preceding claim. It is bookkeeping inside the proof of the full-lift detection proposition.

Sources

PaperOpen
Section
§3.4 Linear lifts and loss of rank
Page
—
Label
—
Line
sections/02-latent.tex:579
Leannot aligned

Phase 2 aligns this node to a Lean declaration.

Friction

  • info
    inferred_dep
    outline: "Expected harmonic rank drop per level" rests on lem:latent-block-kernel
  • info
    inferred_dep
    outline: "Expected harmonic rank drop per level" rests on claim-prop-latent-full-lift-detection-1