University of Pennsylvania:
PhD student in computer science,
BS in computer science and mathematics (summa cum laude),
Verifying Autonomous Systems:
with Rajeev Alur, Osbert Bastani, and Shaan Vaidya
Neural networks are widely used as autonomous agents,
but they are computationally expensive to verify.
We plan to train an automous agent that
will be efficiently verifiable.
Using an F1/10 car to test our approach,
we intend to use
Imitation-Projected Programmatic Reinforcement Learning
to achieve our goal of a safe self-driving F1/10 car.
Verifying eBPF Programs:
with Santosh Nagarakatte and Srinivas Narayana
I wrote a translator
that takes in eBPF programs written in an assembly-like syntax
and outputs a Z3 query specifying the relationship
between its inputs and outputs.
This project studies the Linux kernel's eBPF verifier with the goal
of proving its soundness (or finding a counterexample).
Professors Nagarakatte and Narayana are continuing the project
with a Facebook Networking Systems Research Award.
Verifying Assembly Implementations of Cryptographic Algorithms:
September 2017–May 2018
with Jay Lim, Santosh Nagarakatte, and Mihai Andrei