The Fisher Paradox: Dissipation Interference in Information-Regularized Gradient Flows
Michael Farmer, Abhinav Kochar, and Yugyung Lee

TL;DR
This paper uncovers a new interference mechanism in Fisher-regularized Wasserstein gradient flows, revealing a paradoxical behavior where dissipation temporarily opposes energy descent, with detailed analysis and simulations confirming the phenomenon.
Contribution
It introduces the Fisher Paradox, analyzes the dynamics on the Gaussian manifold, and provides exact solutions and simulations demonstrating the interference effect in information-regularized gradient flows.
Findings
Identification of a positive cross-dissipation sign below a critical scale
Exact Riccati-type variance equation with three dynamical regimes
Confirmation of the effect beyond Gaussian assumptions through simulations
Abstract
We show that Fisher-regularized Wasserstein gradient flows exhibit a previously unrecognized interference mechanism in their dissipation identity: a cross-dissipation term whose sign becomes positive when the state width falls below a critical scale. In this regime the geometric Fisher channel transiently opposes descent of the baseline free-energy functional, producing what we term the Fisher Paradox. Restricting the flow to the Gaussian manifold yields an exact Riccati-type variance equation with a closed-form trajectory, exposing three dynamical regimes separated by two critical scales: sigma = 1 (cross-dissipation sign flip) and sigma = sqrt(epsilon) (Fisher takeover). The variance potential V(u) = u^2 - 2u - epsilon ln(u) contains a logarithmic centrifugal barrier that shifts the equilibrium attractor by Delta sigma approx epsilon/4. The interference persists for a duration t_cross…
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.
Taxonomy
TopicsAdvanced Thermodynamics and Statistical Mechanics · Statistical Mechanics and Entropy · Geometric Analysis and Curvature Flows
