Undergraduate Research Opportunities in Harvard Groups

These are interesting research opportuniities in Harvard groups, but they do not have funding attached.  (Earlier in the term, I would have recommended using these as the basis for proposals through the URAF office -- e.g., HCRP, which closed in late March.)
Newsletter Sent Monday, April 8, 2019
Harvard - Cambridge, MA
Research Opportunity in Professor Stuart Shieber's group working on interpretability of neural network models for language
Neural networks have made significant improvements in many natural language processing (NLP) tasks. However, they are difficult to interpret and are also susceptible to hidden biases. We are working on improving our understanding of various neural network models for NLP. As part of this project, we’re looking for students to contribute to our work on removing biases from datasets with neural networks. This project is a great opportunity to apply machine learning skills in novel areas. 
Preferred background: CS 181/2 or equivalent. 
Contact information: send resume and relevant course list to Yonatan Belinkov (belinkov@seas.harvard.edu



If you are interested in checking out research opportunities, I encourage you to watch this video link taken from the Summer Research & Job Search presentation on Tuesday, November 13, 2018.  Greg Llacer who heads the Undergraduate Research and Fellowships Office (URAF) discusses the major funding programs available to students.  This includes the PRISE program, for which you want to start preparing before you leave campus this term.  This 37-minute video is a good introduction, and there several longer URAF meetings the week after Thanksgiving (Undergraduate Research Week) that you can attend.  The URAF website has a lot of information about resources available to undergraduates wanting to pursue research.  I also have an extensive list of links for Internships at universities (e.g., REU's), national labs, and other research organizations.  These provide a wide range of experiences, which may fall closer to individual research or, alternately, closer to those of industrial internships. 

Perhaps the biggest part of setting up a research experience for an undergrad -- and the main reason you need to start early -- is that almost all of these are done under the mentorship and guidance of a professor.  The first step here is to identify a professor whose research interests you.  In SEAS, there is a handy collection of faculty websites with each field broken down into several topics.  (CS even has a short form of this - a google spreadsheet with professors office hours.)  If you are looking outside SEAS, check the department website for lists of faculty.  Spend some time on these websites and pick out a few professors who have research that interests you. 

The next step is that you need to contact the professors.  For most, this means emailing them.  (A few have said that students should just knock on their doors, but that is not the norm.)  Realize that professors get a lot of emails, including many from students outside Harvard that are looking for internships.  So spend some time on the email message.  Don't send out a "form letter" to several professors.  They can usually figure that out, and it is a good way to get the unread message deleted. It is probably a good idea to copy the professor's assistant.  Each professor is unique, but there are some common threads that I have heard.  They want students who have a genuine interest in doing research, so you should make the case as to why you are interested in their particular research.  Another theme that I've heard is that students often don't realize the amount of time this will take.  It varies, but figure 10-12 hours per week during the school year.  Think carefully about how you will do this with your other commitments.  If you are applying for summer funding, they may want you to do some work during the school year.  Attach a resume or CV, including any past relevant experience and courses.