Skip to Main Content
An artificial brain project in which conceptual models are specified algorithmically is discussed. The feature that distinguishes the approach used in this project from connectionist neural models is its vertical abstraction of brain structure and function. The models comprise multiple local networks of neurons, each of which is a complex network of molecular and chemical processes. The memory manipulation system and reference neuron mechanism used in this approach are described. The development of simulation models for the different conceptual components and the assessment of these models are reviewed.