I'm interested in multi-task and transfer learning, particularly for multilingual models, and I'm trying to find efficient ways to use and train large text encoder models like BERT. Recently I did an internship at Facebook AI and mainly worked on using pre-trained models for machine translation. I've previously worked on approximate inference (e.g. variational inference, MCMC, ABC), and am excited about Bayesian deep learning, and more recently improving the calibration of neural networks.