{"id":"arxiv/2609.35716","url":"https://arxiv.org/abs/2609.35716","tags":["math.NA"],"score":0,"title":"When does least-squares residual minimization solve a differential equation?","domain":"math.NA","source":"arxiv","authors":["Carlos Esteve-Yagüe","Richard Tsai","Stanley Osher"],"reviews":0,"subject":"math.NA","summary":"Least-squares residual minimization approximates a differential equation through an optimization problem. This paper investigates the conditions required to transition from stationary points and minimizing sequences of the residual functional to the exact solution. Starting from the classical orthogonality relation between the residual and the range of the linearized operator, we study how boundary conditions and parameterized variations determine what can be inferred about the residual at a stationary point. Smooth spurious critical points can occur even for well-posed problems; for linear…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-09-28","reactions":{}}