A Universal Trade-off Between the Model Size, Test Loss, and Training Loss of Linear Predictors.

Published in: CoRR (2022)

Keyphrases