Rafael Miyazaki, Cosmin Pohoata, Michael Zheng
Around 10 years ago, Agol and Krushkal showed that the number of chromatic polynomials $P_{G}$ arising from graphs $G$ on $n$ vertices grows exponentially with $n$, by establishing that the (dual) flow polynomial $F_{G}\left(\frac{3+\sqrt{5}}{2}\right)$ already takes on exponentially many values, if one varies $G$ over all planar cubic graphs $G$ on $n$ vertices. We show, more generally, that the size of the set $\{P_G(q): |V(G)|=n\}$ is exponential in $n$, for every fixed real number $q \neq 0,1,2$. In fact, our approach can also be pushed to show that $P_{G}(q)$ already takes on exponentially many values, if we only vary $G$ over all planar graphs on $n$ vertices. The case $q=3$ confirms a conjecture of Agol, which was initially motivated by the $\mathsf{NP}$-completeness of planar $3$-colorability.
Michael Zheng
This paper is about the rainbow dual of the Hales Jewett number, providing general bounds an anti-Hales Jewett Number for hypercubes of length k and dimension n denoted $ah(k, n).$ The best general bounds this paper provides are: $(k-1)^n < ah(k, n) \leq \frac{(k-1)^2-2}{k-1}\cdot k^{n-1}+\frac{k+1}{k-1}.$ This paper also includes proofs about the specific cases of $k = 2$ and $k = 3$, where we show that $ah(2, n) = 2$ and $2^n < ah(3, n) \leq 3^{n-1} - 2\cdot3^{n-4} + 2$ for all natural numbers n $>$ 4. For $n < 4$, we have found the exact values: $ah(3, 1) = 3$, $ah(3, 2) = 5$, and $ah(3, 3) = 11$. In the case $n = 4$, we have found that $23 < ah(3, 4) \leq 27$.
Maria Axenovich, Michael Zheng
A proper edge-coloring of a graph is an interval coloring if the labels on the edges incident to any vertex form an interval of consecutive integers. Interval thickness s(G) of a graph G is the smallest number of interval colorable graphs edge-decomposing G. We prove that s(G)=o(n) for any graph G on n vertices. This improves the previously known bound of 2n/5 by Asratian, Casselgren, and Petrosyan. While we do not have a single example of a graph with interval thickness strictly greater than 2, we construct bipartite graphs whose interval spectrum has arbitrarily many arbitrarily large gaps. Here, an interval spectrum of a graph is the set of all integers t such that the graph has an interval coloring using t colors. Interval colorings of bipartite graphs naturally correspond to no-wait schedules, say for parent-teacher conferences, where a conversation between any teacher and any parent lasts the same amount of time. Our results imply that any such conference with $n$ participants can be coordinated in o(n) no-wait periods. In addition, we show that for any integers t and T, t<T, there is a set of pairs of parents and teachers wanting to talk to each other, such that any no-wait schedules are unstable -- they could last t hours and could last T hours, but there is no possible no-wait schedule lasting x hours if t<x<T.
Michael Zheng
We establish an upper bound on the minimum codegree necessary for the existence of spanning, fractional Steiner triple systems in $3$-uniform hypergraphs. This improves upon a result by Lee in 2023. In particular, together with results from Lee's paper, our results imply that if $n$ is sufficiently large and satisfies some necessary divisibility conditions, then a $3$-uniform, $n$-vertex hypergraph $H$ contains a Steiner triple system if every pair of vertices forms an edge in $H$ with at least $0.8579n$ other vertices.
Anton Molnar, Cosmin Pohoata, Michael Zheng, Daniel G. Zhu
We show that the Kneser graph of triangulations of a convex $n$-gon has chromatic number $n-2$.
Anton Molnar, Cosmin Pohoata, Michael Zheng
For every $n \geq 5$, we show that the Kneser graph of triangulations of a convex $n$-gon contains a Hamiltonian cycle.
Alexander Clifton, Hong Liu, Letícia Mattos, Michael Zheng
For a graph $G$ and a hereditary property $\mathcal{P}$, let $\text{ex}(G,\mathcal{P})$ denote the maximum number of edges of a subgraph of $G$ that belongs to $\mathcal{P}$. We prove that for every non-trivial hereditary property $\mathcal{P}$ such that $L \notin \mathcal{P}$ for some bipartite graph $L$ and for every fixed $p \in (0,1)$ we have \[\text{ex}(G(n,p),\mathcal{P}) \le n^{2-\varepsilon}\] with high probability, for some constant $\varepsilon = \varepsilon(\mathcal{P})>0$. This answers a question of Alon, Krivelevich and Samotij.