LLMs can construct powerful representations and streamline sample-efficient supervised learning

Add code
Mar 12, 2026

Share this with someone who'll enjoy it:

View paper onarxiv icon

Share this with someone who'll enjoy it: