About

I think curiosity is what makes humans feel alive. It’s what makes us explore, discover, invent, try things, get them wrong, and come back a little smarter. I’d like to see AI models pick up that same drive, and I think it’s part of what real intelligence will look like. That’s the idea behind CuriosLM, the family of small language models I’m building.

It’s also how I learn, and that’s what my blog posts are about: I train small language models from scratch on a tight compute budget (tens to a few hundred dollars per study), building on open-source work. I change one piece at a time and write up what happens, including what didn’t work.

Background: I trained as a physicist (PhD in computational physics), worked briefly in quantum computing research, and am currently an Applied Scientist at AWS.

Views are my own and don’t represent my employer.