Research Lines

From fast surrogates, to interpretable identified dynamics, to understanding the physics inside generative models.
Physics in Generative Models
Probing and steering what video generation models internally represent about physics.
Interpretable System Identification
Learning sparse, physically structured, and uncertainty-aware dynamics in latent spaces.
Interpretable System Identification
Data-Driven Surrogate Modeling
Fast, non-intrusive reduced-order models that replace expensive simulations of structural and biomechanical systems.
Data-Driven Surrogate Modeling