Reproduce “Port-Hamiltonian Neural Networks for Systems with Multiple Asymptotically Stable Equilibria” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Simon Heilig, Jens Püttschneider, Mohammad Itani, Asja Fischer, Timm Faulwasser
The authors introduce ms-PHNN, a port-Hamiltonian neural network with multiple minima, to model dissipative systems with several stable equilibria. It forms energy from products of Bregman-divergence factors, one per target equilibrium, while retaining energy-based dynamics. On Duffing, it captures attractors and achieves derivative RMSE 0.0009 ± 0.0007, versus 0.0039 for NODE; on Four-Magnets, its test RMSE is worse than NODE's.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.