Antagonistic interactions can stabilise fixed points in heterogeneous linear dynamical systems

Samuel Cure,Izaak Neri

DOI: https://doi.org/10.21468/SciPostPhys.14.5.093

2024-05-27

Abstract:We analyse the stability of large, linear dynamical systems of variables that interact through a fully connected random matrix and have inhomogeneous growth rates. We show that in the absence of correlations between the coupling strengths, a system with interactions is always less stable than a system without interactions. Contrarily to the uncorrelated case, interactions that are antagonistic, i.e., characterised by negative correlations, can stabilise linear dynamical systems. In particular, when the strength of the interactions is not too strong, systems with antagonistic interactions are more stable than systems without interactions. These results are obtained with an exact theory for the spectral properties of fully connected random matrices with diagonal disorder.

Disordered Systems and Neural Networks

What problem does this paper attempt to address?

The paper attempts to address the problem of how antagonistic interactions stabilize fixed points in heterogeneous linear dynamical systems. Specifically, the study focuses on whether antagonistic interactions (i.e., negative correlations) can stabilize linear dynamical systems in a fully connected random matrix model with non-uniform growth rates. When the interactions are not too strong, systems with antagonistic interactions are more stable than those without interactions. This conclusion is derived by analyzing the spectral properties of fully connected random matrices, and in the absence of correlations between interaction strengths, systems with interactions are always less stable than those without interactions. The paper primarily validates this finding through theoretical analysis and numerical simulations, demonstrating how antagonistic interactions affect the stability of the system.