2403.10405
Action Functional as an Early Warning Indicator in the Space of Probability Measures via Schrödinger Bridge
Peng Zhang, Ting Gao, Jin Guo, Jinqiao Duan
incompletemedium confidence
- Category
- Not specified
- Journal tier
- Specialist/Solid
- Processed
- Sep 28, 2025, 12:56 AM
- arXiv Links
- Abstract ↗PDF ↗
Audit review
The paper defines the early-warning indicator I(ρ*_t) as 1/t times the minimal Benamou–Brenier kinetic action to connect (ρ0, ρ*_t) (its Eq. (25)), and motivates using Schrödinger bridges to obtain the path marginals ρ*_t, but it does not prove the time-rescaling identity that yields I(ρ*_t)=W2^2(ρ0,ρ*_t)/t^2, nor does it supply theoretical conditions ensuring existence/regularity of the SB interpolation or quantitative tipping guarantees beyond experiments. The candidate solution supplies these missing ingredients: a correct time-rescaling proof for the Benamou–Brenier action on [0,t], standard hypotheses for SB existence, and an explicit Gaussian/Brownian example that produces a discontinuous jump in I. Minor notational issues in the paper (e.g., writing W2 instead of W2^2 in the dynamic formulation) reinforce the need for a rigorous reconciliation. Taken together, the paper’s claims are conceptually sound but theoretically incomplete, while the model’s solution is correct and fills the gaps (indicator definition: ; SB–BB background: ; tipping definition: ).
Referee report (LaTeX)
\textbf{Recommendation:} major revisions \textbf{Journal Tier:} specialist/solid \textbf{Justification:} The paper advances a promising and timely idea: framing early-warning indicators directly in probability space via Schrödinger bridges. The background exposition and empirical studies are engaging. However, the theoretical core is incomplete: the indicator is defined but its fundamental time-rescaling identity and closed-form relation to W2\^2/t\^2 are not established; well-posedness assumptions for the SB interpolation are unstated; and no quantitative tipping guarantee is provided. Addressing these points and correcting minor notational issues will substantially improve rigor and impact.