longform

diverse preference learning — ICLR 2025

improving output diversity during alignment posttraining

Read More

on the creation of narrow AI — NeurIPS 2025

building capable but narrow AI systems

Read More

appreciation

a framing i find useful

Read More

sophomore fall — course reviews and happenings

reflection on fall classes and other general happenings

Read More

freshman year — course reviews and happenings

reflection on freshman year — mostly classes, but i also have a life

Read More

quick takes

burnout as aversive triggers

September 1, 2025

alignment targets

December 20, 2024