In the window of the 2025-07-24 arXiv submission, the Hi-pi method suggests a shift in how machine learning handles high-dimensional physical systems. While symbolic regression has long sought to map data patterns to equations, the sheer volume of redundant dimensionless parameters in complex systems often prevents the discovery of physically meaningful descriptions. The approach, titled “Hierarchical Dimensionless Learning (Hi-pi): A physics-data hybrid-driven approach for discovering dimensio