{"id":"arxiv/2610.27888","url":"https://arxiv.org/abs/2610.27888","tags":["cs.AI","math.NA"],"score":0,"title":"Automatic Differentiation with Constraints in Forward and Inverse Problems","domain":"cs.AI","source":"arxiv","authors":["Nikolay Yavich","Alexander Ryabov","Viacheslav Naumov","Sayan Ranu","N. M. Anoop Krishnan","Evgeny Burnaev","Vladimir Vanovskiy"],"reviews":0,"subject":"cs.AI","summary":"Approximate solutions of partial differential equations (PDEs) on unstructured grids can be represented on a graph. Numerical optimization involving stiff evolutionary and static PDEs often requires solving a large system of equations, which can be treated as a constraint. In this work, we present a fully differentiable graph-based implicit solver that incorporates this constraint. We evaluate its performance across four key tasks: a) linear and nonlinear forward modeling of evolutionary and static problems, b) solving inverse problems to estimate equation coefficients, c) solving inverse…","category":"research","comments":0,"disputed":false,"division":"cs","mentions":0,"discussed":0,"published":"2026-10-20","reactions":{}}