GABRIELE PERUGINI

GABRIELE PERUGINI

Courses a.y. 2024/2025

Biographical note

I graduated in Physics at the Rome University “la Sapienza” in 2013 and received a PhD in Physics from the same University in 2017. During my PhD I worked on the statistical Physics of disordered systems and optimization problems. In 2018 I joined Bocconi University as a PostDoc doing research in machine learning and artificial neural networks. Since 2024 I am a Lecturer in the Computing Science Department.


Research interests

I am interested in the statistical mechanics description of learning and optimization problems and in understanding to what extent the knowledge of simple learning models can help predict the phenomenology of deep neural networks. 
Recently I focused on the empirical and analytical description of the loss landscape of artificial neural networks. 
 


Selected Publications

Annesi, Brandon Livio; Lauditi, Clarissa; Lucibello, Carlo; Malatesta, Enrico M.; Perugini, Gabriele; Pittorino, Fabrizio; Saglietti, Luca
Star-shaped space of solutions of the spherical negative perceptron
PHYSICAL REVIEW LETTERS, 2023

Negri, Matteo; Lauditi, Clarissa; Perugini, Gabriele; Lucibello, Carlo; Malatesta, Enrico
Storage and learning phase transitions in the random-features Hopfield model
PHYSICAL REVIEW LETTERS, 2023

Baldassi, Carlo; Malatesta, Enrico M.; Perugini, Gabriele; Zecchina, Riccardo
Typical and atypical solutions in nonconvex neural networks with discrete and continuous weights
PHYSICAL REVIEW. E, 2023

Baldassi, Carlo; Lauditi, Clarissa; Malatesta, Enrico M.; Perugini, Gabriele; Zecchina, Riccardo
Unveiling the Structure of Wide Flat Minima in Neural Networks
PHYSICAL REVIEW LETTERS, 2021

Lucibello, Carlo; Pittorino, Fabrizio; Perugini, Gabriele; Zecchina, Riccardo
Deep learning via message passing algorithms based on belief propagation
MACHINE LEARNING: SCIENCE AND TECHNOLOGY, 2022