{"id":"arxiv/2606.21236","url":"https://arxiv.org/abs/2606.21236","tags":["physics.comp-ph","math.NA","math.OC","physics.flu-dyn"],"score":0,"title":"Physics-Informed Neural Networks for coupled stiff transport systems","domain":"physics.comp-ph","source":"arxiv","authors":["Laetitia Laguzet","Gabriel Turinici"],"reviews":0,"subject":"physics.comp-ph","summary":"Purpose: Physics-Informed Neural Networks (PINNs) struggle with stiff, regime-changing transport equations due to instability, loss imbalance, and violations of physical consistency. This paper investigates these failures through the Marshak wave equations - a canonical benchmark from radiative transport - where initial and boundary conditions differ by up to 12 orders of magnitude, and proposes targeted modifications to the standard PINN framework to overcome them. Design/methodology/approach: Three modifications are introduced: (1) a ScaledSigmoid final activation enforcing physical bounds…","category":"research","comments":0,"disputed":false,"division":"physics","mentions":0,"discussed":0,"published":"2026-06-19","reactions":{}}