Sidharth Baskaran

About

I am currently a founding engineer and researcher at Automorphic (YC S23) in San Francisco, conducting research on building steerable, knowledgeable language models. I'm also affiliated with Confirm Labs, working on intepretability projects such as a hypernetwork-based model editor and new ways to leverage SAEs for circuit discovery.

My general research interests include intepretability, alignment, domain adaptation methods, and foundation models for science.

For the past year I've been on leave from Georgia Tech, where I finished my first year studying Computer Science.

I also enjoy rowing my own gears, cycling, music of the past, and a good book or blog.

I have a resume, and you can reach me via email.