Generative models for (anti)neutrino-nucleus scattering
Date:
Abstract: We will present our recent results on the development of AI-driven models for simulating (anti)neutrino–nucleus collisions. We focus on (anti)neutrino energies characteristic of accelerator-based neutrino experiments, such as T2K. In this energy range, the theoretical description of (anti)neutrino–nucleus interactions is complex, with nuclear effects playing a crucial role. We propose AI-driven models capable of learning from both theoretical predictions and experimental data. In particular, we will discuss generative models, such as GANs, NF, and VAEs, and assess their ability to generalize and extrapolate data-encoded information into kinematic regions that are not yet accessible experimentally.