• Corpus ID: 239998146

Learning Graph Cellular Automata

  title={Learning Graph Cellular Automata},
  author={Daniele Grattarola and Lorenzo Francesco Livi and Cesare Alippi},
Cellular automata (CA) are a class of computational models that exhibit rich dynamics emerging from the local interaction of cells arranged in a regular lattice. In this work we focus on a generalised version of typical CA, called graph cellular automata (GCA), in which the lattice structure is replaced by an arbitrary graph. In particular, we extend previous work that used convolutional neural networks to learn the transition rule of conventional CA and we use graph neural networks to learn a… 

Figures and Tables from this paper

Variational Neural Cellular Automata

This work proposes a generative model, the Variational Neural Cellular Automata (VNCA), which is loosely inspired by the biological processes of cellular growth and differentiation, and shows that the VNCA can learn a purely self-organizing generative process of data.

Attention-based Neural Cellular Automata

This work presents a new class of attention-based NCAs formed using a spatially localized—yet globally organized—self-attention scheme named Vision Transformer Cellular Automata (ViTCA), yielding superior performance across all benchmarks and for nearly every evaluation metric.

Learning stochastic dynamics and predicting emergent behavior using transformers

We show that a neural network originally designed for language processing can learn the dynamical rules of a stochastic system by observation of a single dynamical trajectory of the system, and can

AMPNet: Attention as Message Passing for Graph Neural Networks

This work proposes AMPNet, a general Graph Neural Network (GNN) architecture for uncovering feature-level interactions between different nodes in a graph, and applies a multiheaded attention operation during message-passing to contextualize messages based on the feature interactions betweenDifferent nodes.

Low-Shot Learning and Pattern Separation using Cellular Automata Integrated CNNs

Experiments under low-shot conditions demonstrate that the cellular automata-integrated CNN outperforms compact state-of-the-art CNN models by 6-10% on static image datasets and 8-12% on temporal image sequence datasets.

Collective intelligence for deep learning: A survey of recent developments

A historical context of neural network research’s involvement with complex systems is provided, and several active areas in modern deep learning research that incorporate the principles of collective intelligence to advance its capabilities are highlighted.

Evolving modular soft robots without explicit inter-module communication using local self-attention

VSRs, aggregations of mechanically identical elastic blocks, are focused on, where the same neural controller is used inside each voxel, but without any inter-voxel communication, hence enabling ideal conditions for modularity: modules are all equal and interchangeable.


This work outlines a form of neural network collectives (NNC), motivated by recent work in the field of collective intelligence, and gives details about the specific sub-components that an NNC may have.



Cellular Automata on Graphs: Topological Properties of ER Graphs Evolved towards Low-Entropy Dynamics

This work extends the investigation towards graphs obtained in a simulated-evolution procedure, starting from Erdő s–Renyi (ER) graphs and selecting for low entropies of the CA dynamics, finding a strong association of low Shannon entropy with a broadening of the graph’s degree distribution.

Evolving Self-organizing Cellular Automata Based on Neural Network Genotypes

The application of a genotypical template for all cells in the automaton greatly reduces the search space for the evolutionary algorithm, which makes the presented morphogenetic approach a promising and innovative method for overcoming the complexity limits of evolutionary design approaches.

Cellular automata as convolutional neural networks

  • W. Gilpin
  • Computer Science
    Physical review. E
  • 2019
This work shows that any CA may readily be represented using a convolutional neural network with a network-in-network architecture, and investigates how the trained networks internally represent the CA rules using an information-theoretic technique based on distributions of layer activation patterns.

A survey of cellular automata: types, dynamics, non-uniformity and applications

This survey tours to the various types of CAs introduced till date, the different characterization tools, the global behavior ofCAs, like universality, reversibility, dynamics etc, and special attention is given to non-uniformity in CAs and especially to non -uniform elementary CAs, which have been very useful in solving several real-life problems.

Statistical mechanics of cellular automata

Analysis is given of ''elementary'' cellular automata consisting of a sequence of sites with values 0 or 1 on a line, with each site evolving deterministically in discrete time steps according to p definite rules involving the values of its nearest neighbors.

Generalization over different cellular automata rules learned by a deep feed-forward neural network

To test generalization ability of a class of deep neural networks, a large number of different rule sets for 2-D cellular automata (CA) are generated based on John Conway’s Game of Life to show generalization to rule sets and neighborhood sizes that were not seen during the training at all.

CA-NEAT: Evolved Compositional Pattern Producing Networks for Cellular Automata Morphogenesis and Replication

This paper proposes a new principle of morphogenesis based on compositional pattern producing networks (CPPNs), an abstraction of development that has been able to produce complex structural motifs without local interactions to provide a valuable mapping for morphogenetic systems, beyond CA systems, where development through local interactions is desired.

Outer-totalistic cellular automata on graphs

Learning Cellular Automation Dynamics with Neural Networks

We have trained networks of Σ - II units with short-range connections to simulate simple cellular automata that exhibit complex or chaotic behaviour. Three levels of learning are possible (in

Computation at the edge of chaos: Phase transitions and emergent computation