Leticia Mattos Da Silva

Ph.D. Candidate in Computer Science

Profile picture of Leticia Mattos Da Silva
Contact
leticiam[at]mit[dot]edu

About me

I’m a Ph.D. candidate at MIT, where I research numerical algorithms for nonlinear PDE and other associated dynamical problems, under the supervision of Justin Solomon. The goal of my work is to advance applications of these PDE in the fields of geometry processing and computer graphics.

My research is supported by a 2024 MathWorks Engineering Fellowship. In the past, I had the generous support of a 2023 MathWorks EECS Fellowship, a 2022 Schwarzman College of Computing Fellowship funded by Google and a 2021 MIT Distinguished Fellowship in EECS.

Before coming to MIT, I graduated cum laude with a bachelor’s degree in mathematics at the University of California, Los Angeles.

Research Interests

My research is focused on numerical methods for PDE, in particular nonlinear parabolic PDE and certain kinetic equations. I am also interested in algorithms to solve dynamical problems involving the stochastic counterparts to these nonlinear PDE, e.g., algorithms for the Schrö​dinger Bridge Problem. I'm mainly concerned with applications in geometry processing, computer graphics, and vision.

Publications

2025

Gummy bears

Variational Elastodynamic Simulation

Leticia Mattos Da Silva, Silvia Sellán, Natalia Pacheco‐Tallaj, Justin Solomon
ACM SIGGRAPH Conference Proceedings
PDF | Code (coming soon)

Resampling S shape

Mirror Bridges Between Probability Measures

Leticia Mattos Da Silva, Silvia Sellán, Francisco Vargas, Justin Solomon
Under Peer Review
PDF

2024

Fokker-Plank on the sphere

A Framework for Solving Parabolic Partial Differential Equations on Discrete Domains

Leticia Mattos Da Silva, Oded Stein, Justin Solomon
ACM Transactions on Graphics (presented at SIGGRAPH 2024)
PDF

2022

Fracture modes

Breaking Good: Fracture Modes for Realtime Destruction

Silvia Sellán, Jack Luong, Leticia Mattos Da Silva, Aravind Ramakrishnan, Yuchuan Yang, Alec Jacobson
ACM Transactions on Graphics
PDF | Code

In The Press

Press article thumbnail: squishy simulation
Animation technique simulates the motion of squishy objects

By Adam Zewe
June 6, 2025

Press article thumbnail: parabolic PDE
A framework for solving parabolic partial differential equations

By Alex Shipps
August 28, 2024

Also featured on the MIT-IBM Watson AI Lab Newsletter

Press article thumbnail: EECS Graduate Women’s Summit
EECS Graduate Women’s Annual Research Summit Increases Research Visibility and Strengthens Community

By Rachel Yang
November 30, 2021

UROP Information

Ph.D. students like me typically are the day‐to‐day mentors for UROPs in our lab. If you are a student interested in doing a UROP with me, please read this page for more information before emailing me.

Availability

Fall 2025: Not available.

Spring 2026: Contact me via email (before Dec. 2025).

Potential UROP Project

I'm our lab's "local expert on parabolic PDE, Schrö​dinger bridges, and other problems in the dynamical universes." We can jointly figure out a research direction in this area, but here's an example of a UROP project we could work on together — it is quite open‐ended!

Overview. The Schrö​dinger Bridge Problem (SBP) seeks to steer an initial probability distribution of a linear stochastic system to a terminal distribution using minimal energy. In this project, we will explore a generalization of the SBP with nonlinear prior dynamics. The goal is to design a numerical solver for a pair of PDE associated with this generalization of the problem. We will then investigate using this solver in a pipeline to recover the optimal variables for the SBP. Here is a good reference to start with, if you are interested in pursuing a project in this direction.

Prerequisites. There are no requirements beyond understanding of linear algebra, but 6.8410 can be helpful.