: It is designed to mimic the structure of standard PyTorch models, allowing users to define a model with simple parameters like width , grid (spline resolution), and k (spline order).
Can achieve higher accuracy with fewer parameters in low-dimensional scientific problems. Example Usage kan.py
: The library includes specific tools for "symbolic regression," where the model attempts to simplify learned splines into exact mathematical formulas (e.g., turning a learned curve into x2x squared : It is designed to mimic the structure
The pykan repository, maintained by the original researchers, provides the tools to build, train, and visualize these networks. grid (spline resolution)
While more parameter-efficient for some tasks, the current implementation is often slower than optimized MLPs.