Asher Labovich is a graduating senior in Applied Math (with Honors) and International and Public Affairs. Asher has spent his time at Brown exploring applications of mathematics and machine learning, starting a machine-learning-driven election prediction and analysis platform called 24cast with the Brown Political Review, and completing a senior thesis on the mathematical underpinnings of looped transformers. Asher is grateful for all the passionate, smart, kind people he has met during his time at Brown. He will be starting at Liquid AI in Boston as a research engineer following graduation.
Fields of Interest
I think we're at the very start of a long list of theoretical advances in machine learning and I'm excited to play a part in them! I'm particularly interested in Looped Transformers and Supervised ML, especially tree-based algorithms and theoretical analyses on inductive biases. Taking DATA 1030 got me interested in tree-based algorithms, and I got into looped transformers a while later when I started thinking about the maximum possible performance of small models with new architectures.
How has your Concentration and Data Fluency Certificate impacted your time at Brown and your future path?
The flexibility of APMA gave me the ability to develop a broad base in mathematics, which has been immensely useful in my research. I was able to take classes not only in APMA but also in DATA, MATH, ECON, and CS, all of which taught me new skills that have remained useful in future years. However, the best part of APMA is the people! Staying after class to finish proofs, working on amazing projects with passionate people, and learning from people with unique and fascinating skillsets made my experience truly special.
What experiences have shaped your time at Brown?
I really enjoyed working on 24cast, a machine-learning-driven election prediction platform hosted by the Brown Political Review, throughout my sophomore and junior year! It taught me how to lead and motivate a team, especially one with diverse interests across the academic spectrum. I'll be working in a similar team startup environment when I graduate and I can't wait to have some of the 24cast vibes once again!
I also loved studying abroad and absolutely recommend that everybody who can do so takes the opportunity. I had the time to travel both solo and in groups across Australia and Asia and they were some of the best months of my life.
Which advisors or mentors have inspired you throughout your degree?
I truly cannot thank Professors Zsom and Peyam enough for their support throughout my years as a TA and researcher.
I was the head TA for Professor Zsom's class, DATA 1030, and he taught me so much about teaching, even giving me the opportunity to give a lecture on tree-based algorithms to the class. DATA 1030 has grown a lot since I took it as a sophomore, and I'm so happy I got to play even a small part in that. For anybody who can take the course, I highly recommend you do so.
Professor Peyam was an immense help throughout my thesis as my project went through many iterations of failure after failure. He helped to keep me on track and gave me a unique perspective on the dynamical stability aspect of my paper that ended up forming a major portion of the paper.
What is next for you?
I'm working at Liquid AI in Boston! I'm a research engineer focusing on novel language model architectures for small, energy-efficient, private models. I finished my thesis on looped transformers a little while ago but have so many more ideas I didn't have time to dive deep into, so I'm psyched to continue working on them! Plus, I love Boston :)
What has been your favorite part of being a student at Brown?
The people! All of my friends are such passionate, smart, kind people and I love nothing more than listening to them talk about their interests. I've never once felt embarrassed or weird to be somebody who loves my work, since all my friends know exactly what that feels like. I am sad to be leaving them but I can't wait to visit!
What advice would you give to the students who come after you?
I think it's really easy to "utilize the open curriculum" as a STEM student by taking a few humanities classes as an aside and then just taking a bunch of required classes within your major. My advice would be to see the open curriculum as just as useful within your major as it is outside of it. Unlike a lot of other schools we have really lax requirements even within our major and I think it's a waste not to use that to your advantage and get a very broad base across disciplines of CS and mathematics. As an example, my thesis -- which I expected to be mostly empirical -- ended up requiring lots of linear algebra, differential topology, and dynamical systems, and while I had the first (go MATH 540!) I had to do a lot of research to learn the latter two. You'll never know what you'll need to finish a problem and it never hurts to have a broad base.