Light-Weight Diffusion Multiplier and Uncertainty Quantification for Fourier Neural Operators
This work addresses reliability issues in scientific and engineering applications by offering a more scalable and uncertainty-aware operator learning method, though it is incremental as it builds on existing FNO foundations.
The paper tackles the scalability and lack of native uncertainty quantification in Fourier Neural Operators for solving partial differential equations by introducing DINOZAUR, a diffusion-based neural operator that reduces parameter count and memory footprint while achieving competitive or superior performance on PDE benchmarks and providing efficient uncertainty quantification.
Operator learning is a powerful paradigm for solving partial differential equations, with Fourier Neural Operators serving as a widely adopted foundation. However, FNOs face significant scalability challenges due to overparameterization and offer no native uncertainty quantification -- a key requirement for reliable scientific and engineering applications. Instead, neural operators rely on post hoc UQ methods that ignore geometric inductive biases. In this work, we introduce DINOZAUR: a diffusion-based neural operator parametrization with uncertainty quantification. Inspired by the structure of the heat kernel, DINOZAUR replaces the dense tensor multiplier in FNOs with a dimensionality-independent diffusion multiplier that has a single learnable time parameter per channel, drastically reducing parameter count and memory footprint without compromising predictive performance. By defining priors over those time parameters, we cast DINOZAUR as a Bayesian neural operator to yield spatially correlated outputs and calibrated uncertainty estimates. Our method achieves competitive or superior performance across several PDE benchmarks while providing efficient uncertainty quantification.