Get our free extension to see links to code for papers anywhere online!

Chrome logo  Add to Chrome

Firefox logo Add to Firefox

Limitations of the NTK for Understanding Generalization in Deep Learning


Jun 20, 2022
Nikhil Vyas, Yamini Bansal, Preetum Nakkiran


   Access Paper or Ask Questions

  • Share via Twitter
  • Share via Facebook
  • Share via LinkedIn
  • Share via Whatsapp
  • Share via Messenger
  • Share via Email

Data Scaling Laws in NMT: The Effect of Noise and Architecture


Feb 04, 2022
Yamini Bansal, Behrooz Ghorbani, Ankush Garg, Biao Zhang, Maxim Krikun, Colin Cherry, Behnam Neyshabur, Orhan Firat


   Access Paper or Ask Questions

  • Share via Twitter
  • Share via Facebook
  • Share via LinkedIn
  • Share via Whatsapp
  • Share via Messenger
  • Share via Email

Revisiting Model Stitching to Compare Neural Representations


Jun 14, 2021
Yamini Bansal, Preetum Nakkiran, Boaz Barak


   Access Paper or Ask Questions

  • Share via Twitter
  • Share via Facebook
  • Share via LinkedIn
  • Share via Whatsapp
  • Share via Messenger
  • Share via Email

Improving the Reconstruction of Disentangled Representation Learners via Multi-Stage Modelling


Oct 25, 2020
Akash Srivastava, Yamini Bansal, Yukun Ding, Cole Hurwitz, Kai Xu, Bernhard Egger, Prasanna Sattigeri, Josh Tenenbaum, David D. Cox, Dan Gutfreund


   Access Paper or Ask Questions

  • Share via Twitter
  • Share via Facebook
  • Share via LinkedIn
  • Share via Whatsapp
  • Share via Messenger
  • Share via Email

For self-supervised learning, Rationality implies generalization, provably


Oct 16, 2020
Yamini Bansal, Gal Kaplun, Boaz Barak


   Access Paper or Ask Questions

  • Share via Twitter
  • Share via Facebook
  • Share via LinkedIn
  • Share via Whatsapp
  • Share via Messenger
  • Share via Email

Distributional Generalization: A New Kind of Generalization


Sep 17, 2020
Preetum Nakkiran, Yamini Bansal

* PN, YB co-first authors 

   Access Paper or Ask Questions

  • Share via Twitter
  • Share via Facebook
  • Share via LinkedIn
  • Share via Whatsapp
  • Share via Messenger
  • Share via Email

Deep Double Descent: Where Bigger Models and More Data Hurt


Dec 04, 2019
Preetum Nakkiran, Gal Kaplun, Yamini Bansal, Tristan Yang, Boaz Barak, Ilya Sutskever

* G.K. and Y.B. contributed equally 

   Access Paper or Ask Questions

  • Share via Twitter
  • Share via Facebook
  • Share via LinkedIn
  • Share via Whatsapp
  • Share via Messenger
  • Share via Email

Minnorm training: an algorithm for training over-parameterized deep neural networks


Jun 21, 2018
Yamini Bansal, Madhu Advani, David D Cox, Andrew M Saxe


   Access Paper or Ask Questions

  • Share via Twitter
  • Share via Facebook
  • Share via LinkedIn
  • Share via Whatsapp
  • Share via Messenger
  • Share via Email