If your inquiry is addressed there, I might prioritized other emails in my evergrowing inbox.

Can I be your TA? [Click to expand!]

Please apply here: I typically do not respond to email updates seeking TA-ship or asking me to review your application because I have all necessary information in our portal. I actually don’t have a final say in who is assigned to me, but so far, my preferences have been taken into account.

Please do not come to my office asking for TA-ship. Appreciate your understanding.

You’re a University of Utah student interested in research opportunities [Click to expand!]

I work with a small group of PhD, MS, and BS students at any given moment, and I meet with everyone regularly. Thus, each semester I typically welcome very few new MS/BS students, and sometimes none at all.

I try to involve NLP graduate students in co-mentoring and I sometimes pair students on the same project. This has been successful in the past since it’s easier to get unstuck with someone else and by doing this more students get to do research with me.

Typically, the students I select to join my lab are those with whom I’ve already established rapport; for example, students that have taken a course I thought, attended lectures and asked insightful questions, and completed a solid project that showcased their potential for research. This means I never work with MS students in the first semester of their masters. I teach:

  • CS 6340/5340 (Natural Language Processing)
  • CS 6966/5966 (Local Explanations for Deep Learning Models)

Please note that if you are already balancing work and study, or if you are engaged in another research project, I believe it may not be realistic for you to take on an additional project. In this case, I would not be able to involve you in my group to ensure that you are not overextended and that everyone’s time is respected.

Expectations: Meeting regularly, making consistent progress, staying organized, and communicating clearly. Progress doesn’t mean that every week you present new successful results and everything works smoothly. Research is rarely like that! Progress involves a weekly cycle where you formulate well-thought-out hypotheses, implement your current solution, gather results, and thoroughly analyze these outcomes. Implementation can take longer than expected due to a sequence of trial and error. In our meetings, you should present a meaningful interpretation of the outcomes, and ideally come up with suggestions for next steps.

If you reach out, please send me an email with the following information: [Click to expand!]
  1. What’s your educational status? [undergrad, masters, PhD, other + which semester/year] Note that I’m not taking MS students in the first semester of their masters.
  2. Which research questions or problems interests you? Why is this lab the right place to conduct this research?
  3. What do you hope to get out of this collaboration?
  4. Are you familiar with PyTorch?
  5. Have you TA-ed for any courses in KSoC? If so, for which courses/instructor, and if you didn’t please make a note of that.
  6. Have you worked with other KSoC faculty? If so, summarize what you worked on, and if you didn’t please make a note of that.
  7. Which courses from the following list have you completed and with which grade:

    • CS 6966 – Local Explanations for Deep Learning Models
    • CS 6340 – Natural Language Processing
    • CS 6353 – Deep Learning
    • CS 6350 – Machine Learning
    • CS 6540 – Human-Computer Interaction
    • CS 6960 – Human-AI Alignment

You are a BS/MS student. Am I going to pay you? [Click to expand!]

BS/MS students I’ve worked so far sign up for an independent study to get class credits. I very, very rarely fund BS/MS students as RAs, and in all cases I initiated the conversation about this.

BS students: If things are going well and there is continued mutual interest in working together, I’m open to helping you with the UROP proposal or supervising your undergraduate thesis.

You are not a student at the University of Utah, but you want to be [Click to expand!]

If you’re interested in doing a PhD in the School of Computing, please apply. We will carefully consider every application. More information about the application process can be found here.

While I’m honored to be considered as anyone’s advisor, as most professors I don’t have bandwidth to answer every email that inform me about achivements and interest to work with me. In very rare cases, when an email is specific and demonstrates that the writer genuinely engaged with my work, I might respond. If I haven’t responded to you, not only that sending more emails won’t help, it actually overwhelms me, and I’d really appreciate if you don’t do that.

Note on the Statement of Purpose [Click to expand!]

Through the years I noticed that many students believe they should focus on maximizing the number of publications to improve their PhD application. While demonstrating research experience is indeed very important, it is not all that matters. I strongly recommend reading “Inside Ph.D. admissions: What readers look for in a Statement of Purpose” by Nathan Schneider.

You can have a great record, but you also must demonstrate good focus and fit. A statement with well-written focus and fit shows developed research taste, knowledge of currently most prominent approaches in the area of interest and the gaps that need to be filled in to make short- and long-term progress, ideas of how to address these gaps, why working with some advisor in some school will help you tackle these questions, etc. Just as publishing, some of these are acquired skills that we do not expect that you already fully mastered when applying for PhD. That’s what a PhD is for. 🙂 Your statement is your chance to demonstrate to a potential advisor, who doesn’t know you yet, that you figured this out to some extent. A great record without showing any of these does not make an application I’d be excited about.

I hope you can infer now how even research projects and activities that did not result in a publication can be useful in your statement. I linked some resources for how to improve these skills here, and you can find examples of great statements here.

You have general questions about admissions [Click to expand!]

Please refer to the Kahlert School of Computing’s Admissions webpage, and the Admissions FAQ. I’d encourage you to also explore the Office of International Admissions website. These pages explain the admissions deadline, requirements, and process.

If you have further questions, please reach out to about the program and to about admissions.

You’re looking for postdoc positions [Click to expand!]

I’m not hiring postdocs yet.

You’re looking for internship positions [Click to expand!]

I don’t have any internship opportunities to offer.

You’re a high school student [Click to expand!]

I don’t work with high school students in any capacity.

How to pronounce your name? [Click to expand!]

My name is pronounced as Ah-nah Mara-so-veetch, with “Mara” as the actress “Mara Wilson”.

Talk bio [Click to expand!]

Ana Marasović is an Assistant Professor in the Kahlert School of Computing at the University of Utah. Her primary research interests are at the confluence of NLP, explainable AI, and multimodality. She aims to rigorously validate AI technologies and make human interaction with AI more intuitive. She was a Young Investigator at the Allen Institute for AI from 2019–2022. During that time, she also had a courtesy appointment in the Paul G. Allen School of Computer Science & Engineering at the University of Washington. She obtained her PhD in 2019 from Heidelberg University. She received Best Paper Award ar ACL 2023, Best Paper Honorable Mention at ACL 2020, and Best Paper Award at SoCal 2022 NLP Symposium.