Step-Controlled DPO: Leveraging Stepwise Error for Enhanced Mathematical Reasoning

Add code
Jul 02, 2024

Share this with someone who'll enjoy it:

View paper onarxiv icon

Share this with someone who'll enjoy it: