Ben Burns | Research

I am broadly interested in the mathematics of generative modeling and in AI4Science. I am especially interested in metrics which detect overfitting in generative models for scientific applications, and how to train generative models to generalize well in high-risk scientific domains given relatively few data examples.

Click here for CV (updated October 2024).

research experience

• Undergraduate Honors Thesis
  • Supervised by Professor Markos Katsoulakis and Dr. Benjamin Zhang
  • Title: A Mean-Field Games Approach to Score-based Generative Modeling
• DARoS Laboratory (2022 - 2024)
  • Supervised by Professor Donghyun Kim
  • Studying robot state estimation and invariant EKFs through Lie groups
• UMass Math REU (2023)
  • Supervised by Professor Markos Katsoulakis and Dr. Benjamin Zhang
  • Studied score-based generative modeling with stochastic differential equations and mean-field games
  • Implemented Hamilton-Jacobi-Bellman regularizer for potential implicit score matching
• UMass Math REU (2022)
  • Supervised by Professor R. İnanç Baykur
  • Studied Stein fillings and mapping class groups, specifically relations of Dehn twists on genus 0 surfaces
  • Wrote an expository paper on fundamental group presentations under local curve conjugation
• Noblis, intern (2022, 2021)
  • Supervised by Dr. Scott James
  • Summary of work available on request