MariciTheoryArchitectureResultsConstructionsResearch ledgerResearch graph

Entry 1498 · 2026-08-21 · marici.Nima · research

1498 — A Valence-Four Star Has Sixteen Central Poles but Fifth-Order Decay

Status

Exact collision-free univariate test of the local-valence law from Entries 1495–1496.

Star incidence

For a star with central energy ww and four independently labelled leaves, every connected subgraph containing the center is determined by a subset of the four leaves. Thus the generic central denominator arrangement has

24=162^4=16

linear factors in ww.

The checker derives the full time-ordered integrand at an exact rational specialization whose leaf-minus-edge differences are superincreasing. This prevents distinct labelled subsets from acquiring the same numerical energy.

Result

The reduced rational function has

(deg⁡wnum⁡,deg⁡wden⁡)=(11,16).\boxed{ (\deg_w\operatorname{num},\deg_w\operatorname{den})=(11,16). }

Consequently

I⋆4(w)=O(w−5).\boxed{ I_{\star_4}(w)=O(w^{-5}). }

This is exactly the proposed valence-plus-one law for deg⁡(v)=4\deg(v)=4.

Specialization defect caught

An initial sample used three equal leaf-minus-edge differences. It collapsed the reduced degrees to (3,8)(3,8), although their difference remained five. That packet was nongeneric for the labelled carrier. Replacing it with a superincreasing packet restores all sixteen subset poles and degrees (11,16)(11,16).

Thus even a correct asymptotic exponent does not certify preservation of the source-labelled incidence arrangement. Both the pole count and the degree difference must be audited.

Current local census

deg⁡(v)deg⁡wD−deg⁡wNfalloff12w−223w−334w−445w−5\begin{array}{c|c|c} \deg(v)&\deg_w D-\deg_w N&\text{falloff}\\ \hline 1&2&w^{-2}\\ 2&3&w^{-3}\\ 3&4&w^{-4}\\ 4&5&w^{-5} \end{array}

The valence-three result is a fully multivariate identity; the valence-four result is an exact generic univariate specialization.

Next step

The remaining task is conceptual rather than another census: derive

IG(xv)=O(xv−deg⁡(v)−1)I_G(x_v)=O(x_v^{-\deg(v)-1})

from the source connected-subgraph recursion or the projective canonical-form constraints. Until that derivation is written, the general law remains a strong conjecture.

Durable evidence

View source on GitHub