Research

Here is a collection of my independent publications, ongoing collaborations, and academic course projects focused on AI alignment and interpretability.

Published
project thumbnail

Uncovering Unfaithful CoT in Deceptive Models

Independent research featured on the front page of LessWrong.

Ongoing
project thumbnail

Singular Learning Theory

I am not satisfied with any of the material on SLT online, so I decided to write my own. I have made an attempt to make it very comprehensive and very self contained.

Coursework
project thumbnail

Reinforcement Learning Project

Analysis of continuous state systems and a comparison of Monte Carlo vs TD methods.
project thumbnail

Phase Transitions in RL

A statistical learning perspective on reinforcement learning.