{"id":"arxiv/2608.11020","url":"https://arxiv.org/abs/2608.11020","tags":["cs.LG","math.NA","physics.comp-ph"],"score":0,"title":"Derivative Computation in PINNs: Automatic Differentiation, Finite Differences and Beyond","domain":"cs.LG","source":"arxiv","authors":["Maciej J. Mikulski","Tadeusz Uhl"],"reviews":0,"subject":"cs.LG","summary":"We systematically investigate finite-difference (FD) derivative computation in Physics-Informed Neural Networks (PINNs) as an alternative to automatic differentiation (AD). On three benchmark PDEs we show that, with a properly calibrated step size, FD matches AD in accuracy on every problem while running faster across the full tested batch-size range and using substantially less GPU memory, and that a stochastic variant we propose outperforms AD on a stationary problem. We further show that for neural architectures with inter-sample dependencies (e.g. BatchNorm, self-attention) the standard…","category":"research","comments":0,"disputed":false,"division":"cs","mentions":0,"discussed":0,"published":"2026-08-11","reactions":{}}