Berfin Şimşek

Berfin Şimşek

fellow @ NYU CDS
guest @ Flatiron CCM

About me

Hi there! I am a Faculty Fellow at New York University (CDS) and a guest researcher at Flatiron (CCM). My research is centered around developing mathematical theories for modern machine learning. Notably, I introduced a novel method to study the loss landscape complexity of neural nets. These days I am analyzing end-to-end learning dynamics in exciting models and interpreting the learning process as well as the resulting weights. I am also co-teaching Machine Learning Course.

Previously, I was a doctoral researcher at EPFL in Computer and Communication Sciences, supervised by Clément Hongler and by Wulfram Gerstner. During my Ph.D, I did a research internship at Meta AI.

Before my Ph.D., I studied Electrical-Electronics Engineering and Mathematics double-major at Koç University. I also earned two bronze medals in International Mathematical Olympiad (IMO). I am an active Argentine Tango dancer and I like Yoga.

See my Google Scholar page for an up to date list of publications.

Publications

Quickly discover relevant content by filtering publications.
Deep Linear Networks Dynamics: Low-Rank Biases Induced by Initialization Scale and L2 Regularization
Geometry of the Loss Landscape in Overparameterized Neural Networks: Symmetries and Invariances