Cartesian genetic programming (CGP), introduced in (Miller and Thomson, 2000), is a form of GP in which programs are represented as an indexed graph and encoded via a linear array of integers (Miller and Thomson, 2000).

CGP has been applied to the evolution of artificial neural networks (ANNs) (Khan et al., 2010), image filters (Harding, 2008), visual object detection in robots (Leitner et al., 2012) and the emergence of multicellular organisms (Rothermich and Miller, 2002).

An example series of game graph domains procedurally evolved through the application of a genotypic structure inspired by CGP.