Pretraining Data Mixtures Enable Narrow Model Selection Capabilities in Transformer Models

Add code
Nov 01, 2023

Share this with someone who'll enjoy it:

View paper onarxiv icon

Share this with someone who'll enjoy it: