Classification of networks of automata by dynamical mean-field theory
Z. Burda, Jerzy Jurkiewicz, Henrik Flyvbjerg · Journal of Physics A Mathematical and General · 1990
Dynamical mean-field theory is used to classify the 65536 different networks of binary automata on a square lattice with nearest-neighbour interactions. Application of mean-field theory gives 700 different mean-field classes, which fall in seven classes of different asymptotic dynamics characterised by fixed points and 2-cycles.