Track new results across pure and applied mathematics.
Search papers, labs, and topics across Lattice.
This paper characterizes the macroscopic phase transition of length-$O(\log N)$ random walk bridges on vertex-transitive Ramanujan graphs with logarithmic girth. Within a sharp $O(1)$ critical window, the walk's maximum displacement abruptly transitions from diffusive $O(\sqrt{\log N})$ to diameter-scale $O(\log N)$, while the normalized range shifts to a non-trivial two-point mixture law. The result pinpoints the precise geometric threshold where constrained diffusion on optimal expanders ceases to behave like a standard Euclidean bridge and begins exploring the full graph volume.
Constrained random walks on Ramanujan expanders do not expand smoothly: their maximum span explodes from $O(\sqrt{\log N})$ to $O(\log N)$ across an ultra-narrow $O(1)$ threshold governed by whether the walk closes in the universal cover.
For vertex-transitive Ramanujan graphs with logarithmic girth, a simple random walk bridge of length of order $\log N$, where $N$ is the size of the graph, has a maximum distance that changes from order $\sqrt{\log N}$ to order $\log N$ in a bounded critical window. We prove this by separating bridges whose lifts to the regular tree close from those whose lifts do not. A uniform two-term return estimate determines the probabilities of these two cases and the real-valued critical center. In the same $O(1)$ window, the normalized range has a two-point limiting law whose mixture weights vary nontrivially across the window.