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

Chrome logo  Add to Chrome

Firefox logo Add to Firefox

Statistical Inference with Stochastic Gradient Algorithms


Jul 25, 2022
Jeffrey Negrea, Jun Yang, Haoyue Feng, Daniel M. Roy, Jonathan H. Huggins

* 41 pgs 

   Access Paper or Ask Questions

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

Understanding Generalization via Leave-One-Out Conditional Mutual Information


Jun 29, 2022
Mahdi Haghifam, Shay Moran, Daniel M. Roy, Gintare Karolina Dziugaite

* 18 pages 

   Access Paper or Ask Questions

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

The Neural Covariance SDE: Shaped Infinite Depth-and-Width Networks at Initialization


Jun 06, 2022
Mufan Bill Li, Mihai Nica, Daniel M. Roy


   Access Paper or Ask Questions

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

Adaptively Exploiting d-Separators with Causal Bandits


Feb 10, 2022
Blair Bilodeau, Linbo Wang, Daniel M. Roy

* 32 pages, 3 figures 

   Access Paper or Ask Questions

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

Towards a Unified Information-Theoretic Framework for Generalization


Nov 17, 2021
Mahdi Haghifam, Gintare Karolina Dziugaite, Shay Moran, Daniel M. Roy

* 22 Pages, NeurIPS 2021, This submission subsumes [arXiv:2011.02970] ("On the Information Complexity of Proper Learners for VC Classes in the Realizable Case") 

   Access Paper or Ask Questions

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

Minimax Optimal Quantile and Semi-Adversarial Regret via Root-Logarithmic Regularizers


Nov 07, 2021
Jeffrey Negrea, Blair Bilodeau, Nicolò Campolongo, Francesco Orabona, Daniel M. Roy

* NeurIPS 2021 
* 30 pages, 2 figures. Jeffrey Negrea and Blair Bilodeau are equal-contribution authors. Updated citations 

   Access Paper or Ask Questions

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

The Future is Log-Gaussian: ResNets and Their Infinite-Depth-and-Width Limit at Initialization


Jun 07, 2021
Mufan Bill Li, Mihai Nica, Daniel M. Roy


   Access Paper or Ask Questions

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

NUQSGD: Provably Communication-efficient Data-parallel SGD via Nonuniform Quantization


May 01, 2021
Ali Ramezani-Kebrya, Fartash Faghri, Ilya Markov, Vitalii Aksenov, Dan Alistarh, Daniel M. Roy

* This entry is redundant and was created in error. See arXiv:1908.06077 for the latest version 

   Access Paper or Ask Questions

  • Share via Twitter
  • Share via Facebook
  • Share via LinkedIn
  • Share via Whatsapp
  • Share via Messenger
  • Share via Email
1
2
3
4
5
6
>>