{"id":"arxiv/2606.09820","url":"https://arxiv.org/abs/2606.09820","tags":["math.FA","cs.LG","math.PR","q-fin.MF","stat.ML"],"score":0,"title":"Weighted universal approximation of differentiable maps on infinite-dimensional manifolds","domain":"math.FA","source":"arxiv","authors":["Philipp Schmocker","Josef Teichmann"],"reviews":0,"subject":"math.FA","summary":"We generalize the universal approximation theorem for functional input neural networks (FNN) to differentiable maps by including the approximation of the derivatives. A FNN maps the input from a possibly infinite-dimensional weighted manifold to the real-valued hidden layer, on which a non-linear scalar activation function is applied, and then returns the output into a Banach space via some linear readouts. By proving a weighted Nachbin theorem, we establish a universal approximation theorem for differentiable maps, which goes beyond the usual formulation on compact sets and also includes…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-06-08","reactions":{}}