Entropy of percolating and random infinite clusters

Cluster entropy, defined in terms of site-valence distribution function, is computed for site percolation on square lattices using a Monte Carlo simulation technique. At the percolation threshold this quantity is found to be insensitive to short-range attractive interactions, where it exceeds the corresponding value given by Hankey in terms of the site occupation probability. Short-range repulsive interaction gives rise to a small enhancement in its value. All of these values are, however, smaller than the entropy of an infinite cluster in which all possible valencies are equally likely to occur. (author)