Inside an exploding supernova, an unimaginable number of neutrinos are born — ghost particles that pass through almost everything without a trace. But when gathered together, they start influencing each other and lightning-fast changing their role — physicists call it 'flavor.' Previously, calculations of these transformations relied on a uniform medium or smoothed-out randomness. But a real supernova is like a dance floor with jerky music: matter density jumps unpredictably, and each such jolt can set the particle's next move.
The authors proposed describing this chaos as a chain dance, where each step depends on the previous one, and a random touch changes the entire motion. Mathematically, this boiled down to simple formulas without hours-long simulations. It turned out that the rhythm of the dance is set by the intermittency of matter: the sharper the density jumps, the more often neutrinos change roles. Whether they heat the star or cool it depends on which of the three neutrino types is the lightest — and that remains a mystery in physics.
as written by the authors arXiv:2601.11272
Fast flavor conversion (FFC) in core-collapse supernovae is usually analyzed in homogeneous backgrounds or with smooth stochastic turbulence closures. We construct an exact linear benchmark in which the matter-noise memory kernel is instead generated by a finite She--Leveque log-Poisson cascade. Projecting the marginal FFC channel onto this kernel gives a causal Volterra equation whose non-Markovian memory closes into a finite local system. The resulting Laplace-space resolvent is rational, with one pole pair for each cascade level, so the dispersion relation, characteristic polynomial, and time-domain solution can be checked analytically. We then connect this benchmark to the realization-level toy model and gain-region heating proxy used in the supplementary derivation. For the updated intermittent choice $$\delta\rho/\langle\rho\rangle=0.4$$, $$\bar\lambda/\mu=1$$, and hence $$\kappa_0=0.16$$, the representative $$N=2$$, $$r=2$$ cascade gives $$\sigma_{\rm int}^2=1.124$$ and an intermittent conversion fraction $$1-P_{\rm base}\simeq0.455$$. The older weaker normalization $$\kappa_0=0.05$$ gives $$1-P_{\rm base}\simeq0.324$$. The corresponding Mori-like heating ratios are $$Q_{\rm int}/Q_{\rm hom}=1.060$$ and $$1.041$$, whereas the Wang/Fornax-like ratios are $$0.855$$ and $$0.899$$. Thus intermittency mainly controls the conversion fraction, while the neutrino spectral hierarchy controls the sign of the heating correction.
Neutrinos are so light and elusive that trillions pass through us every second — and almost nobody notices.