ROCKIN' AI
2-6 September 2024
ROCKIN' AI is continuing in 2025! Visit ROCKIN'AI 2.0!
The expanding scale of machine learning models and datasets, along with methodological innovations, is rapidly extending the capabilities of artificial intelligence. Therefore, addressing fundamental questions and thoroughly understanding the basic building blocks of models and training algorithms is of utmost importance.
Given the high-dimensional nature of real-world data, tools from high-dimensional inference and statistical physics are particularly apt for tackling these theoretical challenges. Since the seminal works on the perceptron by Gardner and Derrida in the 1980s, there has been a rich interplay across disciplines, yielding a plethora of innovative ideas and insights into neural network functioning.
This multidisciplinary conference seeks to unite leading researchers from the fields of statistical physics, mathematical physics, and machine learning. Our goal is to offer diverse perspectives on crucial topics in contemporary machine learning, including:
Associative memories, diffusion models, energy-based models;
Representation learning, structured data modeling;
Language modeling, self-supervised learning, alternative learning paradigms;
Mathematical physics approaches to high-dimensional probability, spin glasses, Boltzmann machines;
Learning dynamics and information propagation in neural networks.
The event is scheduled for September 2nd to 6th, 2024, in Roccella Jonica, Calabria, Italy, an ideal setting for scientific exchange and close to a marvelous beach. Alongside the scientific agenda, we will host a social dinner and organize hiking groups for the free afternoon in Aspromonte, offering stunning views of the Ionian Sea.
Stay tuned for more details!
Deadline for applications: May 15th. Registration is free.
Location: Convento dei Minimi, Largo Molinero, 89047 Roccella Ionica RC. You will see the banner of the conference hanging outside.
The schedule of the event is now available at Scientific programme.
Carlo Lucibello PRIN 2022 CUP: J53D23001330001
Pierluigi Contucci PRIN 2022 CUP: J53D23003690006
Luca Saglietti PRIN 2022 CUP: J53D23001390001
Jean Barbier ERC CHORAL project number: 101039794