Sang Michael Xie
@sangmichaelxie

ML PhD @StanfordAILab advised by Percy Liang and Tengyu Ma.




Sang Michael Xie    @sangmichaelxie
Fine-tuning destroys some pre-trained info. Freezing parameters *preserves* it and *simplifies* the learning problem -> better ID and OOD accuracy. Excited to present Composed Fine-Tuning as a long talk at #ICML2021! Paper: https://t.co/C8xHLiumEA Talk: https://t.co/I2BRAADOqJ

Sang Michael Xie    @sangmichaelxie
WILDS includes some nice examples of real world distribution shifts, including spatial and temporal shifts in satellite image tasks: poverty (PovertyMap) and building/land use prediction (FMoW). Hope these help spur progress in ML for sustainability + social good! #ICML2021











 








Sang Michael Xie

Sergey Levine

DeepMind

Aran Komatsuzaki

Andrew Gordon Wilson

Cambridge MLG