NESM-neuro-evolutionary-system

This repository provides a reference implementation of the Neuro-Evolutionary System Model (NESM), a biologically inspired artificial agent integrating perception, prediction, action selection, and energetic homeostasis within a unified functional architecture.

The code accompanies the manuscript “A Neuro-Evolutionary System Model (NESM): Integrating Perception, Action and Energetic Homeostasis in Artificial Agents” and is intended to support reproducibility of the experimental results reported therein.

The repository includes:

• the core implementation of the NESM architecture (cortex, cerebellum, basal ganglia, thalamus, brainstem, spinal cord, and sensorimotor interfaces),

• configurable training scripts for reinforcement-learning-based experiments,

• multimodal simulation environments,

• diagnostic and evaluation notebooks,

• configuration files defining hyperparameters and experimental settings,

• scripts for logging, visualization, and performance analysis.

The implementation is provided for research and academic purposes and is designed to illustrate the functional and algorithmic principles of the NESM model rather than optimized large-scale deployment.