Advising statement

(back to home)


My students currently work on two broad research themes: generalization in natural language processing (compositionality, systematicity and transfer) and learning from language (constructing and learning from natural language explanations in general machine learning settings). Here's a little about how we operate:

Goals PhD programs are designed both to generate knowledge and to train new researchers. These functions are equally important, but I think it's more valuable for students to explore a broad set of problems, technical tools and collaboration styles than to produce a coherent body of work on a single topic. By the end of their time at MIT, students should have their own research agenda and the expertise to pursue it independently.

Research style We're an NLP group, and value linguistic insight and expertise as much as machine learning expertise. Our current modeling toolkit involves lots of neural networks, though history suggests that this could change at any minute. The fact that we work with language data won't change. We try to avoid low-hanging fruit and tune our baseline models carefully. The function of papers is to generate insight, not fractional improvements in accuracy, and a successful project is one that shapes the way researchers think long after the specific model or task being described goes out of use.

Group structure About half the current students came with previous NLP experience and the other half had other ML backgrounds. The group is currently five PhD students, some co-advised, along with a rotating group of visitors, undergrads and master's students. Individual interests range from databases to cognitive science. I expect the group to consist of six to eight people at steady state. I prioritize personal interaction and want to play an active role in every project that my students are working on. I also want to keep the group small enough that everyone knows (and has opinions about!) what everyone else is doing—peer mentoring is often more useful than top-down advising.

Meetings I expect to meet with students one-on-one a minimum of once a week. (Some students also like to do short check-ins much more frequently.) These meetings can involve high-level project planning, low-level technical discussions, and general life / career chat. In addition to one-on-one meetings, we hold (1) a formal meeting once a week which generally features a paper discussion or practice talk, and (2) a casual hangout where people share personal and project updates.

Life outside the lab We like to spend time on things other than research! Work–life balance is important for intellectual development, health and personal relationships. We organize group retreats once a year (this year we're going skiing organizing a socially-distant picnic and some rousing games of Among Us), and many people are also involved in athletic, artistic or cultural pursuits outside the lab. The office is usually empty on weekends.