18th September - Rectorate Hall, Università del Salento
09:00-09:30 Registration
09:30-10:00 Welcome and Institutional Greetings
Prof. Fabio Pollice, Rector of the University of Salento
Prof. Michele Campiti, Head of the Department of Mathematics and Physics "Ennio De Giorgi"
Session 1
Chair: Silvio Franz
10:00-10:30 Sergio Caracciolo, Università Statale di Milano
An atlas for matchings
10:30-11:00 Marc Mézard, Bocconi University in Milano
Are Neural Networks Collision Resistant?
11:00-11:30 Coffee break
11:30-12:00 Ada Altieri, Université Paris Cité
Decoding emergent patterns in large microbial communities through the lens of disordered systems
12:00-12:30 Felix Ritort, Universitat de Barcelona
Molecular Replica Symmetry Breaking
12:30-13:00 Matteo Marsili, ICTP in Trieste
Unifying principles of learning with and without brains
13:00-15:00 Light lunch break
Session 2
Chair: Linda Albanese
15:00-15:30 Riccardo Zecchina, Bocconi University in Milan
Dynamical Learning in Deep Asymmetric Recurrent Neural Networks
15:30-16:00 Simone Warzel, Technical University of Munich
The quantum AT line in the self-overlap corrected SK model
16:00-16:30 Mauro Pastore, ICTP in Trieste
Statistical mechanics of extensive-width Bayesian neural networks near interpolation
16:30-17:00 Antonio Sclocchi, University College London
Probing and Learning the Hidden “Grammar” of Data with Diffusion Models
17:00-18:30 Tea break with a musical interlude
Orchestra popolare “I filì tu spezialu Giorgi”
Istituto Comprensivo Calimera-Caprarica-Martignano
19th September - Great Hall, CNR-Nanotec, Campus Ecotekne
09:00-09:15 Late Registration
09:15-9:30 Welcome and Institutional Greetings
Prof. Fabrizio Illuminati, director of the CNR-Nanotec
Session 3
Chair: Adriano Barra
9:30-10:00 Pierluigi Contucci, Università di Bologna
LLM and Number Theory: a Case Study
10:00-10:30 Simona Cocco, Ecole Normale Supérieure in Paris
Mutational Paths in Sequence Landscapes inferred from Data
10:30-11:00 Coffee break
11:00-11:30 Giorgio Parisi, Sapienza Università di Roma
The lower critical dimension for spin glasses and the behavior of the correlation function in a tube geometry
11:30-12:00 Francesco Zamponi, Sapienza Università di Roma
Neural Network architectures for efficient sampling of statistical physics models
12:00-12:30 Federico Ricci-Tersenghi, Sapienza Università di Roma
Optimizing Spin Glasses with the help of Machine Learning
12:30-14:30 Light Lunch break with visit to CNR-Nanotec Labs
Session 4
Chair: Fabrizio Durante
14:30-15:00 Anita Mehta, Oxford University
Hearings, mishearings and word decryption
15:00-15:30 Aurélien Decelle, Universitad Complutense de Madrid
RBMs, powerful inference machines
15:30-16:00 Enzo Marinari, Sapienza Università di Roma
Spurious overlaps in attractor networks
16:00-16:30 Coffee break
16:30-17:00 Alessandra Lanotte, CNR-Nanotec in Lecce
Inferring the Turbulent Breakup of Colloidal Aggregates Using Graph Neural Networks
17:00-17:30 Yasser Roudi, King’s College University of London
Long-term behaviour of learning machines under closed-loop Learning
20:00 Social Dinner (registration required)
Here you can download the program and book of abstract.