Graph rules for recurrent neural network dynamics: extended version

TL;DR

This paper provides an extended survey on how graph structures influence recurrent neural network dynamics, including new results, derivations, and open questions to guide future research.

Contribution

It offers an expanded overview with additional results and open questions, enhancing understanding of graph-based neural network dynamics.

Findings

Includes new derivations and figures

Provides open questions for future research

Extends previous survey with additional results

Abstract

This is an extended version of our survey article, "Graph rules for recurrent neural network dynamics," to appear in the April 2023 edition of the Notices of the AMS. It includes additional results, derivations, figures, references, and a set of open questions.