Fig. 3
From: Learning versus optimal intervention in random Boolean networks
A diagrammatic representation of the XCS system, adapted from (Karlsen and Moschoyiannis 2018a). The XCS system consists of the entire diagram, excluding the uppermost box which represents the environment that the XCS system interfaces with (in this instance, the Boolean network). The current environment state is in the form of a bit string, such as 10010. The feedback supplied to the feedback mechanism is a reward value (either 0 or 1000 in this instance). The effector instruction consists of simply an index value, which instructs the effector to perform a bit flip on the bit string value at that index