A New Kind of Network? Review and Reference Implementation of Neural Cellular Automata

2026-04-27

AI Summary

Traditional cellular automata struggle to learn complex update rules from data, limiting their applicability in modeling complex systems. This work proposes the first unified theoretical framework and modular symbolic notation for neural cellular automata (NCA) and introduces NCAtorch, an open-source PyTorch implementation. By providing a standardized, reproducible, and extensible reference implementation, this study bridges cellular automata with deep learning, enabling data-driven learning of self-organizing generative behaviors. The resulting framework establishes a new foundation and benchmark platform for modeling complex self-organizing systems.

Abstract

Stephen Wolfram proclaimed in his 2003 seminal work "A New Kind Of Science" that simple recursive programs in the form of Cellular Automata (CA) are a promising approach to replace currently used mathematical formalizations, e.g. differential equations, to improve the modeling of complex systems. Over two decades later, while Cellular Automata have still been waiting for a substantial breakthrough in scientific applications, recent research showed new and promising approaches which combine Wolfram's ideas with learnable Artificial Neural Networks: So-called Neural Cellular Automata (NCA) are able to learn the complex update rules of CA from data samples, allowing them to model complex, self-organizing generative systems. The aim of this paper is to review the existing work on NCA and provide a unified modular framework and notation, as well as a reference implementation in the open-source library NCAtorch.

Problem

• Cellular Automata
• Neural Cellular Automata
• complex systems
• self-organizing systems
• generative modeling

Innovation

• Neural Cellular Automata
• learnable update rules
• self-organizing systems
• modular framework
• NCAtorch