Postdoc on “The Energy Landscape and Dynamics of Machine Learning Algorithms”

Applications are invited to fill one postdoctoral fellowship on “The Energy Landscape and Dynamics of Machine Learning Algorithms”, possibly starting in the Fall 2019 (or earlier). The fellowship is for one year and is funded by the Simons collaboration “Cracking The Glass Problem”.

The successful candidate will work at ENS Paris under the supervision of Giulio Biroli. The goal of the postdoc is to develop theoretical frameworks for Deep Neural Networks and test them in numerical experiments.   She/he will belong to a team based at ENS Paris and EPFL Lausanne lead by Giulio Biroli and Matthieu Wyart working on statistical physics approaches to Deep Learning. 

In order to apply, a short motivation letter, a CV, and two recommendation letters have to be sent at the email address: giulio.biroli@ens.fr

The deadline for applications is January 15, 2019.