Differentiating through Stochastic Differential Equations: A Primer
R. Leburu, L. Nurbekyan, & L. Ruthotto. (2026). "Differentiating through Stochastic Differential Equations: A Primer." arXiv:2601.08594
CV
Education
- B.S. in Applied Mathematics, Emory University, May 2026 (expected)
- GPA: 4.0
Research Experience
- Oak Ridge National Laboratory (Summer 2025)
- Research Intern, Mentor: Dr. Guannan Zhang
- Developed stochastic optimal control approaches using conditional diffusion models to solve Bayesian inverse problems
- Researched hybrid methods combining diffusion posterior sampling with SOC techniques
- Implemented multiple SOC solvers: adjoint matching, finite element method, neural ODEs, neural-FBSDE
- Emory University (Ongoing)
- Undergraduate Researcher with Prof. Lars Ruthotto & Prof. Levon Nurbekyan
- Working on algorithms for high-dimensional stochastic optimal control
- Exploring connections between flow matching, score-based models, mean field games, and control theory
- Emory University REU - Computational Mathematics for Data Science (Summer 2024)
- Undergraduate Researcher, Advisor: Dr. Deepanshu Verma
- Investigated methods to enhance variational autoencoders using conditional normalizing flows
- Built and trained VAE models using Pyro with PyTorch
- Emory University (Ongoing)
- Undergraduate Researcher with Prof. Matthias Chung
- Working on variational sparse paired autoencoders for inverse problems
Publications
Variational Sparse Paired Autoencoders for Inverse Problems
J. Solomon, R. Leburu, M. Li, & M. Chung. (in preparation). "Variational Sparse Paired Autoencoders for Inverse Problems."
Matching Algorithms for High-Dimensional Stochastic Optimal Control
R. Leburu, L. Nurbekyan, L. Ruthotto, & G. Zhang. (in preparation). "Matching Algorithms for High-Dimensional Stochastic Optimal Control."
Presentations
Enhancing Variational Autoencoders with Conditional Normalizing Flows
Poster at Joint Mathematics Meetings (JMM) 2025 Undergraduate Poster Session, Seattle, WA
Technical Skills
- Languages: Python, Java, C#, R, HTML/CSS
- Scientific Computing & Frameworks: PyTorch, Pyro, FEniCS, NumPy, SciPy, Matplotlib
- Tools: Git, MATLAB, VS Code, LaTeX, SLURM
Relevant Coursework
- Graduate: Topics in Computational Mathematics (Generative Modeling), Matrix Analysis, Numerical Analysis, Real Analysis, Numerical PDEs
- Undergraduate: Mathematical Statistics, Nonlinear Optimization, Linear Algebra (proof-based), Differential Equations, PDEs, Data Structures and Algorithms
Awards & Honors
- Chair’s Achievement Award, Emory Math Department (2025)
- Dean’s List, Emory University (2023-2025)