Experience

I'm on the preparedness team, working to keep AI safe on our way to AGI.

Mechanistic estimation methods for predicting neural network behavior more efficiently than sampling.

I learned many things thanks to the kindness of Jacob Hilton and Paul Christiano.

Post-training alignment auditing agents.

Worked on my system 2 thinking.

I'm a math major and a classics minor, class of 2027, currently on leave.

I continued taking Latin, which I started in middle school. In my last Latin class, I wanted to leave something to the field, and it became my first introduction to doing something with AI.

I also did a lot of math, which for now just sits in some Overleaf files. I intend to link them here as they slowly might get published. For now, my favorite two are:

I worked my way through the math PhD qualifying exams in my first two years.

I worked on trying to understand Transformers, through some theory and empirics, and a bit of interpretability on LLMs.

I'm thankful to have been supported by the Goldwater scholarship.

Part-time during school, I was curious what SFT data really looks like, so I worked with Outlier under Scale. I supervised the creation of datasets for math reasoning, and worked on Humanity's Last Exam.

Strichartz estimates of the Schrödinger equation on the 1D torus.

In high school I focused a lot on math, but I also enjoyed working with friends on the robotics team, leading the programming subteam as we made it to the world championship both years.

Since this timeline is more professionally flavored, I'll recount the more academic part of my childhood. In early elementary school, I took courses through the Johns Hopkins Center for Talented Youth. Then I took undergraduate math courses locally at UNC. In middle school, I took graduate classes at NC State, which I continued to do throughout high school.