An Optimisation Framework for the Well-Conditioned Training of Physics-Informed Neural Networks
Physics-informed neural networks (PINNs) have emerged as a promising route to solve partial differential equations, yet they have struggled to reach the precision of classical solvers. The obstacle is increasingly understood to be one of optimisation, owing to the severely ill-conditioned loss landscape. We present $\textbf{DSGNAR}$: Doubly-Sketched Gauss-Newton with Adaptive Ratio, a scalable second-order optimisation framework that confronts this ill-conditioning and, in doing so, obtains unprecedented accuracy and speed. $\textbf{DSGNAR}$ couples a doubly-sketched Gauss-Newton model with a
Lineage graph
Paper → model → repo connections mined from source citations (Tier-1 exact match).
Why these links exist
Every edge carries a method, confidence, and the source snippet that justified it — so bad links are debuggable.
- PossiblePossibly related (embedding) · 57%NVIDIA/physicsnemo →
- PossiblePossibly related (embedding) · 50%Hamiltonian Neural Networks from a Differential Geometry Perspective [D] →
- PossiblePossibly related (embedding) · 48%pytorch/pytorch →
- LinkedLinked via arxiv author · 85%Joseph Webb →
“An Optimisation Framework for the Well-Conditioned Training of Physics-Informed Neural Networks”
- LinkedLinked via arxiv author · 85%Sadok Jerad →
“An Optimisation Framework for the Well-Conditioned Training of Physics-Informed Neural Networks”
- LinkedLinked via arxiv author · 85%Coralia Cartis →
“An Optimisation Framework for the Well-Conditioned Training of Physics-Informed Neural Networks”
- PossiblePossibly related (embedding) · 75%SciML/NeuralPDE.jl →
- PossiblePossibly related (embedding) · 58%Principled approaches for extending neural architectures to function spaces for operator learning →