autoresearch-automl: extended / benchmarked autoresearch with LLM-based & classical HPO

[ferreirafabio]

Hey @karpathy, I benchmarked classical HPO and LLM agents on autoresearch. If you like it, would appreciate a quick reference in your readme.md:

Repo: https://github.com/ferreirafabio/autoresearch-automl

Cheers,
Fabio