The authors derive and test a parametrization for FNOs that keeps optimal hyperparameters stable as the number of Fourier modes grows, enabling zero-shot transfer from small proxy models to near-billion-parameter models.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
Maximal Update Parametrization and Zero-Shot Hyperparameter Transfer for Fourier Neural Operators
The authors derive and test a parametrization for FNOs that keeps optimal hyperparameters stable as the number of Fourier modes grows, enabling zero-shot transfer from small proxy models to near-billion-parameter models.