{"id":"arxiv/2604.22355","url":"https://arxiv.org/abs/2604.22355","tags":["cs.LG","math.OC","stat.ML"],"score":0,"title":"SOC-ICNN: From Polyhedral to Conic Geometry for Learning Convex Surrogate Functions","domain":"cs.LG","source":"arxiv","authors":["Kang Liu","Jianchen Hu","Wei Peng"],"reviews":0,"subject":"cs.LG","summary":"Classical ReLU-based Input Convex Neural Networks (ICNNs) are equivalent to the optimal value functions of Linear Programming (LP). This intrinsic structural equivalence restricts their representational capacity to piecewise-linear polyhedral functions. To overcome this representational bottleneck, we propose the SOC-ICNN, an architecture that generalizes the underlying optimization class from LP to Second-Order Cone Programming (SOCP). By explicitly injecting positive semi-definite curvature and Euclidean norm-based conic primitives, our formulation introduces native smooth curvature into…","category":"research","comments":0,"disputed":false,"division":"cs","mentions":0,"discussed":0,"published":"2026-04-24","reactions":{}}