6 papers
A statistical physics framework for optimal learning
Francesca Mignacco, Francesco Mori
Learning is a complex dynamical process shaped by a range of interconnected decisions. Careful design of hyperparameter schedules for artificial neural networks or efficient alloca…
Theory of Optimal Learning Rate Schedules and Scaling Laws for a Random Feature Model
Blake Bordelon, Francesco Mori
Setting the learning rate (LR) for a deep learning model is a critical part of successful training. Choosing LRs is often done empirically with trial and error. In this work, we ex…
Optimal switching strategies for navigation in stochastic settings
Francesco Mori, L. Mahadevan
When navigating complex environments, animals often combine multiple strategies to mitigate the effects of external disturbances. These modalities often correspond to different sou…
Analytic theory of dropout regularization
Francesco Mori, Francesca Mignacco
Dropout is a regularization technique widely used in training artificial neural networks to mitigate overfitting. It consists of dynamically deactivating subsets of the network dur…
Cosmic-ray transport in inhomogeneous media
Robert J. Ewart, Patrick Reichherzer, Shuzhe Ren +6
A theory of cosmic-ray transport in multi-phase diffusive media is developed, with the specific application to cases in which the cosmic-ray diffusion coefficient has large spatial…
Optimal Protocols for Continual Learning via Statistical Physics and Control Theory
Francesco Mori, Stefano Sarao Mannelli, Francesca Mignacco
Artificial neural networks often struggle with catastrophic forgetting when learning multiple tasks sequentially, as training on new tasks degrades the performance on previously le…