The Structure of Conservative Gradient Fields
Adrian S. Lewis, Tonghua Tian · SIAM Journal on Optimization · 2021
The classical Clarke subdifferential alone is inadequate for understanding automatic differentiation in nonsmooth contexts. Instead, we can sometimes rely on enlarged generalized gradients called “conservative fields,” defined through the natural pathwise chain rule: one application is the convergence analysis of gradient-based deep learning algorithms. In the semialgebraic case, we show that all conservative fields are in fact just Clarke subdifferentials plus normals of manifolds in underlying Whitney stratifications.