Jesse Michel
I am a PhD student in the Programming Systems Group at MIT advised by Professor Michael Carbin. My research focuses on problems in computer graphics, physical simulation, and scientific computing. I approach these problems by designing programming languages with expressive primitives such as integration and with support for automatic differentiation.
I completed my undergraduate at MIT in 2019 with a BS in Pure Mathematics and Computer Science with a minor in Philosophy. I completed my MEng at MIT studying automatic differentiation of arbitrary-precision arithmetic for performance.
I've enjoyed my time interning in industry at companies working on network theory at IBM Research, LLMs at ASAPP Research, search at Google, and machine learning in the cloud at Microsoft. I enjoy maintaining a typed argument parser with my friend Kyle and a wonderful community of open-source contributors. The library has been used widely in the machine learning community and has been downloaded millions of times.
Research
Learning to Compile Programs to Neural Networks
Logan Weber, Jesse Michel, Alex Renda, Michael Carbin
ICML 2024.
Distributions for Compositionally Differentiating Parametric Discontinuities
Jesse Michel, Kevin Mu, Xuanda Yang, Sai Bangaru, Elias Rojas Collins, Gilbert Bernstein, Jonathan Ragan-Kelley, Michael Carbin, Tzu-Mao Li
OOPSLA 2024.
Systematically Differentiating Parametric Discontinuities
Sai Praveen Bangaru*, Jesse Michel*, Kevin Mu, Gilbert Bernstein, Tzu-Mao Li, Jonathan Ragan-Kelley
SIGGRAPH 2021 (Paper)(Website)(Video)
λS: Computable semantics for differentiable programming with higher-order functions and datatypes
Benjamin Sherman, Jesse Michel, Michael Carbin
Principles of Programming Languages 2020 (Paper)
Sensitivities for Guiding Refinement in Arbitrary-Precision Arithmetic
Jesse Michel
Thesis supervised by Ben Sherman and advised by Michael Carbin 2020 (Thesis)
NAP: Noise-Based Sensitivity Analysis for Programs
Jesse Michel*, Sahil Verma*, Benjamin Sherman, Michael Carbin
Workshop on Approximate Computing Across the Stack (WAX) 2019 (Summary)
Sound and Robust Solid Modeling via Exact Real Arithmetic and Continuity
Benjamin Sherman, Jesse Michel, Michael Carbin
International Conference on Functional Programming 2019 (Paper)
Directed Random Geometric Graphs
Jesse Michel*, Sushruth Reddy*, Rikhav Shah*, Sandeep Silwal*, Ramis Movassagh
Journal of Complex Networks 2019 (Paper)
Teaching
Programming Language Design
As a graduate teaching assistant, I helped develop a new class with Jack Feser, Michael Carbin, and Armando Solar-Lezama. My duties involved helping to design course materials, holding office hours, and grading (Course).
Signal Processing
As an undergraduate teaching assistant, I worked under Adam Hartz and Dennis Freeman. My duties involved helping students in office hours, grading, and discussing the course materials in group meetings (Course).
Fundamentals of Programming
I spent six semesters involved with this class working under Adam Hartz. I spent a semester as a student lab assistant, four semesters as a lab assistant, and a semester as a teaching assistant. I designed exam questions, improved the course website CatSoop, graded exams, and helped students in office hours (Course).
Awards
Silver in the graduate student research competition at PLDI 2023Ashar Aziz Presidential Fellowship 2020
Gold in the undergraduate student research competition at PLDI 2019 (Paper)
Distinguished Paper Award at ICFP 2019
First Place Overall at HackMIT 2018 (Ennui)
First Place Overall at HackMIT 2017 (Pixelator)