Hi, I'm Joe. I'm a Research Engineer at Hugging Face, an open source startup that develops state-of-the-art NLP and machine learning tools. As a member of the science team, I mostly work on research projects related to transfer learning and zero/few-shot sequence classification with large pre-trained models.
Before joining Hugging Face, I completed my master's at Harvard in Data Science where I did research projects with Sasha Rush (commonsense and NLP) and with Finale Doshi-Velez (Bayesian Neural Networks). During my stint at Harvard, I interned with IBM Research where I conducted a project on disentanglement in VAEs.