I'm currently at Google Research in New York, focusing on Natural Language Understanding. I recently released my first coauthored paper, Well-Read Students Learn Better: On the Importance of Pre-training Compact Models. Previously, I worked on Federated Learning at Google Research in Seattle.