Tensor PDE model of biological network formation

Jan Haskovec, Peter Markowich, Giulia Pilli

correctmedium confidence

Category: Not specified
Journal tier: Specialist/Solid
Processed: Sep 28, 2025, 12:56 AM
arXiv Links: Abstract ↗PDF ↗

Audit review

The paper proves convexity of ED under conditions (3.1)–(3.2) and constructs unique global gradient-flow solutions (with positivity preserved and the energy inequality) via subdifferential theory on the closed convex cone H1_0,+(Ω) of symmetric positive semidefinite tensors. The candidate solution reaches the same conclusions by a slightly different route: a supremum (envelope/Danskin) representation for the pressure term and a test with the negative part C− to preserve positivity. These approaches agree on the main results. Minor gaps in the candidate solution concern (i) not explicitly restricting the energy domain to positive semidefinite tensors (needed so p[C] is well-posed and the supremum is finite) and (ii) a terse lower-semicontinuity justification for the pressure term; the paper supplies a careful l.s.c. proof. Overall, both are correct, with the model using a different proof style and requiring small clarifications.

Referee report (LaTeX)

\textbf{Recommendation:} minor revisions

\textbf{Journal Tier:} specialist/solid

\textbf{Justification:}

Both the paper and the candidate solution correctly establish convexity, existence/uniqueness of the gradient flow, energy dissipation, and preservation of positive semidefiniteness for the tensorial network-formation PDE. The paper’s proofs are fully rigorous and include a careful lower-semicontinuity argument for the pressure term; the model solution presents a clean alternate route via an envelope/Danskin representation and a test with the negative part but should explicitly restrict the energy domain to the positive cone and reference a rigorous l.s.c. argument. With these minor clarifications, the model proof would match the paper’s level of rigor.