Espen Slettnes MIT, Class of 2027

I am a junior majoring in Mathematics and Artificial Intelligence & Decision Making at MIT. My current interests include AI, machine learning, and algorithms, and I work as a lab assistant for a machine learning class. My passion for mathematics and problem-solving began early, guiding my path through middle and high school where I concurrently enrolled at UC Berkeley, studying mathematics, physics, and economics. During this time, I participated in the MIT PRIMES-USA internship program, researching combinatorics, including game theory, graph theory, and learning algorithms. I also became deeply involved with Berkeley Math Circle, where I have taught collegiate math topics to high school students, designed proof-based monthly contests, and lead a team of high school and college collaborators. When I am not exploring math or coding various projects, you can also find me playtesting video games or social games of my own design on my friends.

I'm broadly interested in applications of mathematical thinking to computational questions in fields such as modelling, reinforcement learning, interpretability for transformers, computer graphics & vision, music, and games. here are some examples of questions i'm trying to answer:

• how differently do different RoPE frequencies in an attention head behave? how does the tokenization scheme affect this behavior?
• is reward enough?
• when will training an LLM on specific tasks improve its performance on unrelated tasks?
• can a neural network be trained to solve sokoban puzzles via reinforcement learning by pitting it against a neural network trained to generate sokoban puzzles? will the generated puzzles be diverse, interesting, and/or difficult for humans?
• what can be inferred about market dynamics from the options market?
• how can we make numerical computation of solutions to elliptic PDEs more efficient when the PDEs have complex boundary conditions?
• what makes an API made to be used in an embedded scripting language good?
• is sokoban decidable if you control two characters simultaneously with the same inputs, but one moves with a 1 second delay?
• can perfect harmonies in 12 TET be achieved with synthetic instruments?

Publications and Preprints:

• Minimal Embedding Dimensions of Rectangle k-Visibility Graphs
  E. Slettnes. Journal of Graph Applications and Algorithms, Vol. 25, no. 1, pp. 59-96, January 2021. DOI: 10.7155/jgaa.00550, ISBN: 1526-1719. Mentor: Dr. Jesse Geneson of San Jose State University.
  
  Explore the 3D models from this paper.
  


• Sharp bounds on the price of bandit feedback for several models of mistake-bounded online learning
  R. Feng, J. Geneson, A. Lee, E. Slettnes, Theoretical Computer Science, Volume 965, 18 July 2023, 113980. DOI: 10.1016/j.tcs.2023.113980. arXiv: 2209.01366
  


• Extracting Tree Statistics from the Quasisymmetric Bernardi Polynomial (March 2020)
  L. Cai, E. Slettnes, and J. Zhou; Mentor: Dr. Duncan Levear of MIT.


• Expected Capture Time and Throttling Number for Cop versus Gambler (May 2019)
  J. Geneson, C. Quines, E. Slettnes, and S. Tsai. arXiv: 1902.05860
  


• Variations of the Cop and Robber Game on Graphs (October 2017)
  E. Slettnes, C. Quines, S. Tsai, and J. Geneson. arXiv: 1710.11352
  

Selected awards and recognitions:

• The Atlas Scholarship, Fellow, 2023.
• National Merit Scholarship, Scholar, 2023.
• Caroline D. Bradley Scholar, Class of 2023.
• USA Mathematical Olympiad Gold Award, 2022.
• Davidson Fellows Honorable Mention, 2021.
• Spirit of Ramanujan Fellow, 2018, 2019, and 2020.
• S. T. Yau High School Science Award Semifinalist, 2019.
• Broadcom MASTERS, First Place in Mathematics, 2018. (Minor Planet 34379 Slettnes was named in my honor.)
• California Science & Engineering Fair, Project of the Year, 2018
• Bay Area Scholastic Writing Award, Silver key in the Personal Essay and Memoir Category, 2022
• The President's Volunteer Service Award, Gold Level, 2021