Development of algorithms and software implementation of the editor of multi-group neurocognitive architectures

Abstract

The research focuses on developing algorithms and software for modeling reasoning and decision-making processes based on multi-agent neurocognitive architectures. The approach is based on an abstraction of brain neuron interactions, where each neuron is considered a rational agent maximizing its objective function. The software is implemented using a client-server architecture: the client is an editor for visualizing and editing agents, while the server is the core that simulates agent interactions and manages their behavior. The architecture supports 3D visualization and network message exchange. Future work aims to improve visualization and optimize the display of a large number of agents.