Failure of moment matching for the multineighbor construction

Let Γn,m,p\Gamma_{n,m,p} and Yn,k,qY_{n,k,q} be the distributions considered in the paper, let zKxz_Kx denote the relevant count associated with a positive kk-pure shape xx, and set p=n1βp=n^{-\frac{1}{\beta}} and q=nkmβq=n^{-\frac{km}{\beta}}. Moment-ratio conjecture. There is a choice of mm, β\beta, and a positive kk-pure shape xx such that the support of Γn,m,p\Gamma_{n,m,p} is the same as that of Yn,kY_{n,k}, but

limnEKΓn,m,p(zKx)EKYn,k,q(zKx)1.\lim_{n\rightarrow\infty}\frac{\mathbb{E}_{K\in\Gamma_{n,m,p}}(z_Kx)}{\mathbb{E}_{K\in Y_{n,k,q}}(z_Kx)}\not=1.

Thus, even when the two distributions have the same support, the associated expected counts need not be asymptotically equal.

Sources & referencesView supporting material

Primary source

Eric Babson and Jan Spaliński, “From Erdos-Renyi graphs to Linial-Meshulam complexes via the multineighbor construction”, arXiv:2309.05149 (2023).

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.