Foundation Model for Conservation Laws / Effective Field Theory for Neural Networks
ABSTRACT
In this talk, a neural operator model specialized for solving conservation laws, the Flux NO model, will be introduced, along with the development of a foundation model that extends it to operate on arbitrary, non-fixed conservation systems. As a separate topic, research on developing neural network theory using the language of effective field theory, a framework originating from quantum field theory, will also be presented.
