How Does Distribution Shift Shape Pretraining Gains in Neural PDE Surrogates?
Pretraining a neural PDE surrogate can reduce the amount of new CFD data needed when geometry or modeled physics changes. However, it remains unclear how different components of distribution shift affect this benefit. We pretrain a surrogate on 254,909 RANS solutions from one airfoil family and fine-tune it on a new family under two target settings with matched freestream range
Reading in Plain English mode — toggle below to switch
Pretraining a neural PDE surrogate can reduce the amount of new CFD data needed when geometry or modeled physics changes. However, it remains unclear how different components of distribution shift affect this benefit. We pretrain a surrogate on 254,909 RANS solutions from one airfoil family and fine-tune it on a new family under two target settings with matched freestream range
This paper highlights the practical benefits and challenges of pretraining for neural PDE surrogates, a critical area for scientific machine learning. Understanding how distribution shift impacts pretraining gains is crucial for efficiently deploying ML models in engineering and scientific domains, reducing the need for extensive new data collection.