All Stories

  1. Self-Auxiliary Distillation for Sample Efficient Learning in Google-Scale Recommenders