{"id":"arxiv/2605.27568","url":"https://arxiv.org/abs/2605.27568","tags":["hep-th","math-ph","nlin.PS"],"score":0,"title":"Solitonic Construction of Artificial Neural Networks from Nonlinear Field Theory","domain":"hep-th","source":"arxiv","authors":["R. A. C. Correa"],"reviews":0,"subject":"hep-th","summary":"We present a field-theoretic construction of a class of artificial neural networks from solitonic degrees of freedom in nonlinear scalar field theory. The purpose is not to rename a standard neural layer in the language of solitons, but to start from a continuum action, restrict the theory to a nonperturbative sector containing localized stable solutions, perform a collective-coordinate reduction, and derive the neural layer as the finite-dimensional input-output map of the reduced solitonic dynamics. In this construction, the computational unit is a projected collective coordinate of a…","category":"research","comments":0,"disputed":false,"division":"physics","mentions":0,"discussed":0,"published":"2026-05-26","reactions":{}}