Preprint
Aug 2026
Neural Renormalization Group Flow for Percolation
To get good performance for two-dimensional site percolation developping a supervised, scale-shared neural architecture, it is key that the learned latent representation exhibits critical fluctuations and scale-dependent flows consistent with the renormalization-group structure of percolation.
Anaclara Alvez, L. Camagna, S. Chibbaro et al.
· 1 citation