| Time |
Speaker |
Title |
| 01/07-11 |
Conference: |
Rough Landscapes: From Physics to Algorithms |
| 1/14, 2:00pm |
All Participants |
Open problems, discussion, brainstorming |
| 1/15, 10:00am |
Juan Pedro Garrahan
Nottingham |
Thermodynamics of trajectories, optimal dynamics and reinforcement learning[Aud][Cam] |
| 1/16, 1:30pm |
Song Mei
Stanford |
Mean field dynamics of two layer/multi layer neural networks[Slides][Podcast][Aud][Cam] |
| 1/17, 10:00am |
Alexander Rakhlin
MIT |
Interpolation |
| 1/18, 1:30pm |
Matthieu Wyart
EPFL |
Discussion: Neural Tangent Kernel and generalisation in overparametrised deep nets[Slides][Podcast][Aud][Cam] |
| 1/21, 2:00pm |
All Participants |
TBA |
| 1/22, 10:00am |
Florent Krzakala
ENS & KITP |
Selected review of 30 years of physics research on artificial neural networks[Podcast][Aud][Cam] |
| 1/23, 1:30pm |
All Participants |
TBA |
| 1/24, 10:00am |
Levent Sagun
EPFL & KITP |
An overview of empirical studies in neural networks[Slides][Podcast][Aud][Cam] |
| 1/25, 1:30pm |
Patrick Charbonneau
(Duke Univ. & KITP)
Grzegorz Szamel
(Colorado State Univ. & KITP) |
Some Steps Towards Cracking the Dynamics of the Glass Problem[Slides][Podcast][Aud][Cam] |
| 01/28 12:15pm |
Giulio Biroli
ENS Paris & KITP |
The many facets of the rough high-dimensional landscape problem[Podcast][Aud][Cam]
KITP Blackboard Lunch |
| 1/28, 2:00pm |
All Participants |
Organizational meet & greet |
| 1/29, 10:00am |
Marco Baity-Jesi
(Columbia Univ. & KITP)
Chiara Cammarota
(King's College London & KITP) |
Activated Dynamics in High-Dimensional Rugged Landscapes[Slides][Podcast][Aud][Cam] |
| 01/30 7:00pm |
Marc Mézard
ENS Paris & KITP |
Artificial intelligence: success, limits, myths and threats
[Podcast][Aud][Cam]
KITP Public Lecture |
| 1/31, 11:00am |
Maciej Koch Janusz
ETH Zurich & KITP |
Information theory, machine learning and real space RG |
| 2/01, 1:30pm |
Roberto Imbuzeiro M. F. de Oliveira
IMPA, Brazil & KITP |
Mean-field limit for certain deep neural networks[Podcast][Aud][Cam] |
| 2/04, 2:00pm |
All Participants |
Organizational meet & greet |
| 2/05, 10:00am |
Carlo Lucibello
Bocconi University & KITP |
Energy Landscapes and Algorithms in Neural Networks[Slides][Podcast][Aud][Cam] |
| 2/06, 1:30pm |
All Participants |
TBA |
| 2/07, 10:00am |
Corey O'Hern
Yale |
The origin of the jamming critical exponents[Slides][Podcast][Aud][Cam] |
| 2/08, 1:30pm |
Henry D. Pfister
Duke University & KITP |
A few experiences with Machine Learning[Podcast][Aud][Cam] |
| 2/11, 1:00pm |
All Participants |
Meet & Greet |
| 2/13, 1:30pm |
Aukosh Jagannath
Harvard |
Dynamics in random landscapes: from spin glasses to inference[Podcast][Aud][Cam] |
| 2/15, 12:00pm |
Andrea Montanari
Stanford & KITP |
Informal considerations on optimality of AMP |
| 2/19, 10:00am |
Jorge Kurchan
ENS Paris & KITP |
Six lessons we learned from glass dynamics |
| 2/20, 1:30pm |
Valentina Ros
ENS Paris & KITP |
Computing the complexity: Kac-Rice and replicas, a comparison[Podcast][Aud][Cam] |
| 2/21, 10:00am |
Antonio Auffinger
Northwestern University & KITP |
Energy landscapes of mean field spin glass models.[Podcast][Aud][Cam] |
| 2/22, 1:30pm |
All Participants |
Brainstorming and Open Problems session on Rough Landscapes and Dynamics |
| 2/25, 2:00pm |
All Participants |
Discussion |
| 2/27, 1:30pm |
Andrey Lokhov
Los Alamos National Laboratory & KITP |
Learning of discrete graphical models[Podcast][Aud][Cam] |
| 2/28, 10:00am |
Sebastian Goldt
IPhT CEA Saclay & KITP |
Generalisation dynamics of online learning in over-parameterised neural networks[Podcast][Aud][Cam] |
| 2/28, 4:00pm |
All Participants |
What's next? Brainstorming and Discussions |
