Bounding the entropy of hard-squares
by Andrew Rechnitzer
Abstract: Start labelling the vertices of the square grid with 0's and 1's with the condition that any pair of neighbouring vertices cannot both be labelled 1. If one considers the 1's to be the centres of small squares (rotated 45 degrees) then one has a picture of square-particles that cannot overlap. This problem of "hard-squares" appears in different areas of mathematics - for example it has appeared separately as a lattice gas in statistical mechanics, as independence sets in combinatorics and as the golden-mean shift in coding theory.
A core question in this model is to quantify the number of legal configurations - the entropy. In this talk I will discuss the what is known about the entropy and describe our recent work finding precise bounds for hard-squares and related problems..
This is work with Yao-ban Chan (U. Montpellier).