Large scale behavior in the Kuramoto-Sivashinsky equation: The Schwinger-Dyson route
arXiv:2607.17915
2026
Architecture
1 ideas extracted · analyzed Aug 30, 2026
What the math gives to ML
The paper uses a Schwinger-Dyson framework to argue that infrared solutions of the Kuramoto-Sivashinsky equation with stable scaling require a negative effective viscosity in the KS convention. The transferable mechanism is a scale-selective competition between low-frequency amplification and high-frequency fourth-order damping, producing a finite unstable band and an ultraviolet cutoff. A neural implementation is a Fourier residual block with a learned second-order gain and fourth-order stabilizer, constrained by a measurable spectral amplification bound.
Ideas from this paper
Unverified
2026
Replace or augment a residual neural layer with a Fourier-domain scale-selective flow containing a learned second-order term and a fourth-order stabilizer. The block permits controlled low-frequency amplification, as required by the KS infrared mechanism, while damping high-frequency feature noise and preventing unbounded spectral growth.
Useful6/10
Difficulty5/10
Novelty7/10