Zi-Hao Shi

TASE-Stabilized Time-Evolving Natural Gradient Methods for Diffusion-Dominated PDEs

The Time-Evolving Natural Gradient (TENG) method evolves neural approximations of time-dependent partial differential equations by projecting time-discrete target states onto the neural-network manifold through local natural-gradient iterations. For stiff diffusion-dominated problems, however, explicit Runge--Kutta met...