jonathanhuml

5 ms·

1. 1.▲
   by jonathanhuml 2y ago

   The pioneers used to ride these babies for miles!
2. 2.▲
   by jonathanhuml 2y ago

   The solution to the L1 regularization problem is actually a specific form of the classical ReLU nonlinearity used in deep learning. I’m not sure if similar results hold for other nonlinearities, but this gave me good intuition for what thre
3. 3.▲
   In Defense of Obnoxious Influencers(jonhuml.substack.com)

   2 points by jonathanhuml 2y ago|0 comments
4. 4.▲
   A New Manhattan Project for Open Source AI(jonhuml.substack.com)

   2 points by jonathanhuml 3y ago|0 comments