{"id":"arxiv/2603.06287","url":"https://arxiv.org/abs/2603.06287","tags":["cs.CE","cs.AI","cs.LG","math.AP"],"score":0,"title":"Learning Where the Physics Is: Probabilistic Adaptive Sampling for Stiff PDEs","domain":"cs.CE","source":"arxiv","authors":["Akshay Govind Srinivasan","Balaji Srinivasan"],"reviews":0,"subject":"cs.CE","summary":"Modeling stiff partial differential equations (PDEs) with sharp gradients remains a significant challenge for scientific machine learning. While Physics-Informed Neural Networks (PINNs) struggle with spectral bias and slow training times, Physics-Informed Extreme Learning Machines (PIELMs) offer a rapid, closed-form linear solution but are fundamentally limited by physics-agnostic, random initialization. We introduce the Gaussian Mixture Model Adaptive PIELM (GMM-PIELM), a probabilistic framework that learns a probability density function representing the ``location of physics'' for…","category":"research","comments":0,"disputed":false,"division":"cs","mentions":0,"discussed":0,"published":"2026-03-06","reactions":{}}