The Cactus Criterion: When Nonlinear Hodge Theory Reduces to Linear on Graphs
Sebastian Pardo-Guerra, Anil Thapa, Jonathan Washburn

TL;DR
This paper introduces a nonlinear selector on graphs based on a specific edge potential, showing it reduces to the linear Hodge projector precisely on cactus graphs, with explicit computations and a Newton method.
Contribution
It characterizes when the nonlinear Hodge-like operator coincides with the linear one on graphs, identifying cactus graphs as the key condition.
Findings
The nonlinear selector is real analytic and matches the Hodge projector to first order.
For cactus graphs, the nonlinear selector equals the linear Hodge projector on all of C^1(G).
A Newton method for computing the nonlinear selector is provided.
Abstract
Let be a finite connected simple graph with a chosen orientation of its edges. For the edge potential we minimize over each affine class . The minimizer is the unique representative satisfying the nonlinear coclosed equation and hence defines a nonlinear selector We show that is real analytic, identify its image as and compute its differential as a weighted Hodge projector. In particular, agrees with the ordinary Hodge projector to first order at the origin, and the first nonlinear correction is cubic. Our main global theorem is a graph-theoretic criterion: for every admissible edge potential -- even, , strictly convex, and non-quadratic -- the associated nonlinear selector…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
