
Logan Engstrom
@logan_engstrom
CS PhD student @ MIT
ID: 1864965229
http://loganengstrom.com 14-09-2013 20:44:36
316 Tweet
1,1K Followers
331 Following

Want state-of-the-art data curation, data poisoning & more? Just do gradient descent! w/ Andrew Ilyas Ben Chen Axel Feldmann Billy Moses Aleksander Madry: we show how to optimize final model loss wrt any continuous variable. Key idea: Metagradients (grads through model training)
