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

Chrome logo  Add to Chrome

Firefox logo Add to Firefox

Second-order regression models exhibit progressive sharpening to the edge of stability


Oct 10, 2022
Atish Agarwala, Fabian Pedregosa, Jeffrey Pennington


   Access Paper or Ask Questions

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

Synergy and Symmetry in Deep Learning: Interactions between the Data, Model, and Inference Algorithm


Jul 11, 2022
Lechao Xiao, Jeffrey Pennington

* Accepted by ICML 2022; 23 pages 

   Access Paper or Ask Questions

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

Wide Bayesian neural networks have a simple weight posterior: theory and accelerated sampling


Jun 15, 2022
Jiri Hron, Roman Novak, Jeffrey Pennington, Jascha Sohl-Dickstein

* ICML 2022 

   Access Paper or Ask Questions

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

Implicit Regularization or Implicit Conditioning? Exact Risk Trajectories of SGD in High Dimensions


Jun 15, 2022
Courtney Paquette, Elliot Paquette, Ben Adlam, Jeffrey Pennington

* arXiv admin note: text overlap with arXiv:2205.07069 

   Access Paper or Ask Questions

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

Precise Learning Curves and Higher-Order Scaling Limits for Dot Product Kernel Regression


May 30, 2022
Lechao Xiao, Jeffrey Pennington

* 32 pages; 4 + 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

Homogenization of SGD in high-dimensions: Exact dynamics and generalization properties


May 14, 2022
Courtney Paquette, Elliot Paquette, Ben Adlam, Jeffrey Pennington


   Access Paper or Ask Questions

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

Covariate Shift in High-Dimensional Random Feature Regression


Nov 16, 2021
Nilesh Tripuraneni, Ben Adlam, Jeffrey Pennington

* 107 pages, 10 figures 

   Access Paper or Ask Questions

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

Understanding Double Descent Requires a Fine-Grained Bias-Variance Decomposition


Nov 04, 2020
Ben Adlam, Jeffrey Pennington

* Published as a conference paper in the Proceedings of the Thirty-fourth Conference on Neural Information Processing Systems; 54 pages; 5 figures. arXiv admin note: text overlap with arXiv:2008.06786 

   Access Paper or Ask Questions

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

Exploring the Uncertainty Properties of Neural Networks' Implicit Priors in the Infinite-Width Limit


Oct 14, 2020
Ben Adlam, Jaehoon Lee, Lechao Xiao, Jeffrey Pennington, Jasper Snoek

* 23 pages, 11 figures 

   Access Paper or Ask Questions

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

Temperature check: theory and practice for training models with softmax-cross-entropy losses


Oct 14, 2020
Atish Agarwala, Jeffrey Pennington, Yann Dauphin, Sam Schoenholz


   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
>>