Language Models can Evaluate Themselves via Probability Discrepancy

Add code
May 17, 2024

Share this with someone who'll enjoy it:

View paper onarxiv icon

Share this with someone who'll enjoy it: