Abstract
We construct and compare three operator learning architectures, DeepONet,Fourier Neural Operator, and Wavelet Neural Operator, in order to learn theoperator mapping a time-dependent applied current to the transmembranepotential of the Hodgkin- Huxley ionic model. The underlying non-linearity ofthe Hodgkin-Huxley dynamical system, the stiffness of its solutions, and thethreshold dynamics depending on the intensity of the applied current, are someof the challenges to address when exploiting artificial neural networks tolearn this class of complex operators. By properly designing these operatorlearning techniques, we demonstrate their ability to effectively address thesechallenges, achieving a relative L2 error as low as 1.4% in learning thesolutions of the Hodgkin-Huxley ionic model.