AI Method Solves Complex Inverse Partial Differential Equations

Researchers at the University of Pennsylvania have developed a new artificial intelligence (AI) method to solve inverse partial differential equations (PDEs). These equations help scientists determine hidden causes from observed effects. The new approach, called "Mollifier Layers," improves how AI handles these complex mathematical problems.

Traditional AI methods struggle with noisy data and require significant computing power for these equations. The Penn team's innovation refines the mathematical process itself. This method smooths data before calculations, making the process more stable and less computationally intensive. The findings were published in *Transactions on Machine Learning Research* and will be presented at the Conference on Neural Information Processing Systems (NeurIPS 2026).

Inverse PDEs are crucial for understanding systems across various scientific fields. They allow researchers to work backward from observed data to uncover underlying forces. For example, they can help infer epigenetic processes from DNA organization. This new AI method could transform fields like genetics, materials science, and fluid dynamics.

One promising application is understanding chromatin, the folded state of DNA inside cell nuclei. Chromatin structures, though tiny, are vital for gene expression. The new AI method can estimate epigenetic reaction rates, which control gene activity. This could lead to predicting how chromatin changes over time and potentially developing new therapies.

The researchers emphasize that the goal is to move from observing complex patterns to understanding the rules that generate them. This deeper understanding could enable scientists to influence and change these systems. The study received support from the National Cancer Institute, the National Science Foundation, the National Institute of Biomedical Imaging and Bioengineering, and the National Institute of General Medical Sciences.