Workshop on Scientific Methods for Understanding Deep Learning

Workshop at the Conference on Neural Information Processing Systems (NeurIPS) 2024

While deep learning continues to achieve impressive results on an ever-growing range of tasks, our understanding of the principles underlying these successes remains largely limited. This problem is usually tackled from a mathematical point of view, aiming to prove rigorous theorems about optimization or generalization errors of standard algorithms, but so far they have been limited to overly-simplified settings. The main goal of this workshop is to promote a complementary approach that is centered on the use of the scientific method, which forms hypotheses and designs controlled experiments to test them. More specifically, it focuses on empirical analyses of deep networks that can validate or falsify existing theories and assumptions, or answer questions about the success or failure of these models. This approach has been largely underexplored, but has great potential to further our understanding of deep learning and to lead to significant progress in both theory and practice. The secondary goal of this workshop is to build a community of researchers, currently scattered in several subfields, around the common goal of understanding deep learning through a scientific lens.

Submission Deadline: Sep 17 ‘24 (AOE).
The workshop will be held on December 15th in Vancouver, Canada.
For latest news about the workshop, follow @scifordl on X/Twitter. Submit questions for our panelists here.


Keynote Speakers

Zico Kolter

Carnegie Mellon University

Hanie Sedghi

Google DeepMind

Yamini Bansal

Google DeepMind

Surya Ganguli

Stanford University

Tom Goldstein

University of Maryland

Panelists

Yasaman Bahri

Google DeepMind

Andrew Gordon Wilson

New York University

Misha Belkin

UC San Diego & Amazon

Eero Simoncelli

New York University

Organizers

Zahra Kadkhodaie

Flatiron Institute & New York University

Florentin Guth

Flatiron Institute & New York University

Sanae Lotfi

New York University

Davis Brown

UPenn & Pacific Northwest National Lab

Micah Goldblum

Columbia University

Valentin De Bortoli

Google DeepMind

Andrew Saxe

UCL


Questions?

Contact us at scienceofdl.workshop.2024@gmail.com or @scifordl.