I read research papers for "fun," try to reimplement them and usually end up debugging my life choices more than my code
- Currently Learning/doing:
- Forensics in Generative modeling - "identifying fake images made by fake models trained on fake data, if your wondering its going great or so I say to myself :("
- Future Plans:
- Mech Interp - aka Mission Impossible: decoding neural nets before they decode me
Quote I live by: If you don't succeed, blame the author of the paper or its just following the laws of modern ML...


