Energy Decay Network (EDeN)
This work addresses the issue of narrow AI for researchers seeking more adaptable and generalizable neural networks, but it appears incremental as it builds on existing genetic and neural principles without clear breakthroughs.
The authors tackled the problem of narrow AI by developing a framework that co-develops neural architecture and unit processes through genetic and real-time influences, aiming to create a diverse and robust network for general tasks with transfer learning capabilities.
This paper and accompanying Python and C++ Framework is the product of the authors perceived problems with narrow (Discrimination based) AI. (Artificial Intelligence) The Framework attempts to develop a genetic transfer of experience through potential structural expressions using a common regulation/exchange value (energy) to create a model whereby neural architecture and all unit processes are co-dependently developed by genetic and real time signal processing influences; successful routes are defined by stability of the spike distribution per epoch which is influenced by genetically encoded morphological development biases.These principles are aimed towards creating a diverse and robust network that is capable of adapting to general tasks by training within a simulation designed for transfer learning to other mediums at scale.