{"id":"arxiv/2601.02951","url":"https://arxiv.org/abs/2601.02951","tags":["math.OC"],"score":0,"title":"Hopfield neural networks as port-Hamiltonian and gradient systems","domain":"math.OC","source":"arxiv","authors":["Arjan van der Schaft"],"reviews":0,"subject":"math.OC","summary":"The structure of continuous Hopfield networks is revisited from a system-theoretic point of view. After adopting a novel electrical network interpretation involving nonlinear capacitors, it is shown that Hopfield networks admit a port-Hamiltonian formulation provided an extra passivity condition is satisfied. Subsequently it is shown that any Hopfield network can be represented as a gradient system, with Riemannian metric given by the inverse of the Hessian matrix of the total energy stored in the nonlinear capacitors. On the other hand, the well-known 'energy' function employed by Hopfield…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-01-06","reactions":{}}