Digital Spikes : Information Representation in ATR ' sCAM - Brain

This paper describes ongoing ATR's CAM-Brain Project, which is an attempt to build large-scale neural networks ('ar-tiicial brains') in a special hardware called "CAM-Brain Ma-chine" (CBM). At the time of writing (March 1998), the project is making eeorts on two fronts-the construction of the CBM, that is scheduled to be operational in the summer of 1998, and attempting to nd an eecient and eeective representation for the binary signaling of ATR's CAM-Brain Machine (CBM), using the so-called "CoDi-1Bit" model. The CBM is an FPGA based hardware accelerator which updates 3D cellular automata (CA) cells at the rate of 100 billion a second, allowing a complete run of a genetic algorithm with tens of thousands of CA based neural net circuit growths and hardware compiled tness evaluations. It is hoped that by using such a device, it will become practical to evolve 10,000s of neural net modules and then assemble them into humanly deened RAM based artiicial brain architectures which can be run by the CBM in real time to control robots, e.g. a robot kitten. Before large numbers of modules can be assembled together, it is essential that the individual modules have a good functionality and evolvability. The "CoDi-1Bit" CA based neural net model uses 1 bit binary signaling, so a representation needs to be chosen based on this fact. This paper discusses the merits and demerits of a representation that we call "Spike Interval Information Coding" (SIIC). The CAM-Brain Project at ATR Labs aims to construct a large-scale brain-like neural network system. If the project succeeds and our expectations are fulllled, these \artiicial brains" will have a large number of potential applications in several diierent elds, from 'smart' domestic appliances to speech processing and robot control. Of course, up to now, this is pure speculation, and we admit there is still a long way to go before we can talk in more concrete terms. However, we believe that to realize a system that possesses a level of functionality and structural complexity similar to real biological brains, the most appropriate way, if not the only way, is to evolve them, as happened in nature. The fundamental approach of the CAM-Brain Project is the growth/evolution of large-scale neural networks. Since the dawn of the Project 1], Cellular Automata (CA) have been chosen as the medium in which to grow the neural networks. CAs meet the requirements of generality and especially scal-ability, …