Local asymptotic conjecture for generalized abundancy coefficients

Let cmathcalHℓ,n(x)=∑k=0nA(ℓ,n,k)xk/n!cmathcal{H}_{\ell,n}(x)=\sum_{k=0}^{n}A(\ell,n,k)x^k/n!, and let λℓ=ℓln⁡(ℓ−1)−ln⁡(Kℓ)\lambda_{\ell}=\ell\ln(\ell-1)-\ln(\mathcal{K}_{\ell}), where Kℓ=Γ(ℓ)ζ(2)⋯ζ(ℓ)\mathcal{K}_{\ell}=\Gamma(\ell)\zeta(2)\cdots\zeta(\ell). For n→∞n\to\infty and k→∞k\to\infty such that k/n(ℓ−1)/ℓk/n^{(\ell-1)/\ell} converges to a quantity in (0,∞)(0,\infty), define λ~ℓ(k/n)=λℓ+εℓ(k/n)\widetilde{\lambda}_{\ell}(k/n)=\lambda_{\ell}+\varepsilon_{\ell}(k/n), where εℓ(t)\varepsilon_{\ell}(t) is a linear combination of terms tj(ln⁡t)rt^j(\ln t)^r with 1≤j≤ℓ−11\leq j\leq\ell-1 and r∈{0,1,… }r\in\{0,1,\dots\}. Main conjecture. Under these conditions, we should have

A(ℓ,n,k)n!=ℓ−12πnekℓ−kℓln⁡(k)+(ℓ−1)kln⁡(n)−λ~ℓ(k/n)k(1+o(1)).\frac{A(\ell,n,k)}{n!}=\frac{\sqrt{\ell-1}}{2\pi n}e^{k\ell-k\ell\ln(k)+(\ell-1)k\ln(n)-\widetilde{\lambda}_{\ell}(k/n)k}(1+o(1)).

The conjecture refines the preceding central limit theorem by incorporating corrections to λℓ\lambda_{\ell} that are needed to obtain order-one accuracy. The paper notes that such corrections can be calculated for Hℓ,n(1)\mathcal{H}_{\ell,n}(1) in related work, but does not establish this local asymptotic formula.

References

Primary source

Shannon Starr, “Some Observations about the "Generalized Abundancy Index"”, arXiv:2505.07051 (2025).

Progress summary

Never refreshed

Nothing recorded yet. Refresh searches the literature and the public web for attempts on this problem, and writes the first summary here.

Solutions 0

No solutions have been posted yet.