arXiv++ Combinatorics

We determine which connected sets of $N$ vertices in the square lattice support the most spanning trees, using all lattice edges between the chosen vertices. Every maximizer minimizes the number of edges joining the set to the rest of the lattice. Its shape is a rectangle with at most two straight strips of vertices removed at opposite corners. For $N\ge18$, the maximizer is unique up to translation, rotation, and reflection: an explicitly determined rectangle with one possibly empty corner strip removed along its longer side. The proof combines comparisons of Laplacian determinants with geometric comparisons of corner deletions. For $N=n^2$, we prove a sharp stability inequality that bounds the symmetric difference from a translated $n\times n$ square in terms of the loss in logarithmic tree count. Thus, nearly optimal tree counts force the vertex set to be close to a square. We also derive an exact product formula and a recurrence for the number of spanning trees in rectangles with a single corner strip removed.

We compare spanning-tree counts at a fixed number of vertices for Cartesian products of paths $P_r$, $r\ge1$, or cycles $C_s$, $s\ge3$. With the number $d$ of factors fixed, an equal-volume tuple that is more balanced in the logarithmic majorization order has at least as many spanning trees, with equality only for permutations. This comparison persists after adjoining any fixed finite connected Cartesian factor; equal-sided blocks therefore uniquely maximize the count at admissible block volume $n^d$. We prove the comparison through strict log-convexity of one-dimensional heat traces in logarithmic length. The Hartman--Watson representation expresses the cycle interpolation as a positive mixture of theta functions, and the path case follows from it. The same heat-trace order decreases every positive inverse-power sum of the nonzero Laplacian eigenvalues, and hence the total effective resistance. Cylinders behave differently: square balancing can fail. For $P_r\square C_s$ of area $N=rs$, extending the product formula to real height and optimizing over integer circumferences $3\le s\le N$ yields an optimum on the scale $N^{1/3}$. Along areas having a divisor asymptotic to that optimum, quantitative estimates place every graph-maximizing circumference on the same scale. For two-dimensional tori, balancing remains optimal, but the gain in the logarithmic spanning-tree count admits no positive uniform lower bound proportional to the decrease in the side sum.

We establish explicit, universal, and distribution-free bounds for the $n$-th cumulant, $κ_n(X)$, of a scalar random variable, controlled solely by an $n$-th order absolute moment functional $M_n(X)$. The bounds take the form $\lvertκ_n(X)\rvert \le C_n M_n(X)$. Our principal contribution is the derivation of coefficients satisfying $C_n \sim (n-1)!/ρ^{\,n}$, which offers an exponential improvement over classical bounds where the coefficients grow superexponentially (on the order of $n^n$). We present a hierarchy of refinements where the rate parameter $ρ$ increases as the functional $M_n(X)$ incorporates more structural information. The most general bound uses the raw moment $M_n(X)=\mathsf{E}[\lvert X\rvert^n]$ with rate $ρ=\ln 2 \approx 0.693$. Using the central moment $M_n(X)=\mathsf{E}[\lvert X-\mathsf{E}[X]\rvert^n]$ improves the rate to $ρ_{\mathrm{cen}} \approx 1.146$, while assuming symmetry yields even higher rates. The proof is elementary, combining the moment-cumulant partition formula with a uniform moment-product inequality. We further prove that while these bounds are strict (not attainable by non-degenerate distributions), they are asymptotically efficient given the limited information of a single moment. The utility of the bounds is demonstrated through applications to analyticity of CGFs and concentration inequalities.