Abstract:We introduce a new information-theoretic measure, fragmentation (F) which can be used to determine how fragmented predictive information is in a system. The concept can be extended to generate fragmentation matrices that can illustrate information flows through digital brains, in the form of directed graphs. Fragmentation and fragmentation matrices can provide new insights into digital brains structure and function, in other words, how causal digital networks "think" and process information. In addition to describing F we demonstrate how it can be used to examine how complex processing arises in neural networks, including differences in lifetime processing and incidents of incidental encryption.
Abstract:Markov Brains are a class of evolvable artificial neural networks (ANN). They differ from conventional ANNs in many aspects, but the key difference is that instead of a layered architecture, with each node performing the same function, Markov Brains are networks built from individual computational components. These computational components interact with each other, receive inputs from sensors, and control motor outputs. The function of the computational components, their connections to each other, as well as connections to sensors and motors are all subject to evolutionary optimization. Here we describe in detail how a Markov Brain works, what techniques can be used to study them, and how they can be evolved.