{"id":"arxiv/2602.03082","url":"https://arxiv.org/abs/2602.03082","tags":["cs.LG","eess.SY","math.OC"],"score":0,"title":"Geometry-Preserving Neural Architectures on Manifolds with Boundary","domain":"cs.LG","source":"arxiv","authors":["Karthik Elamvazhuthi","Shiba Biswal","Kian Rosenblum","Arushi Katyal","Tianli Qu","Grady Ma","Rishi Sonthalia"],"reviews":0,"subject":"cs.LG","summary":"A growing number of neural architectures have been proposed to enforce geometric constraints, including projection-based networks, exponential-map updates, constrained output layers, and manifold neural ODEs. We provide a unified framework for these geometry-preserving architectures by organizing them according to where and how constraints are enforced, either throughout the intermediate layers or only at the final output. This perspective reveals several gaps in the existing theory. To address these gaps, we prove high-level approximation theorems for projected neural ODEs, intermediate…","category":"research","comments":0,"disputed":false,"division":"cs","mentions":0,"discussed":0,"published":"2026-02-03","reactions":{}}