REVS: Unlearning Sensitive Information in Language Models via Rank Editing in the Vocabulary Space

Add code
Jun 13, 2024

Share this with someone who'll enjoy it:

View paper onarxiv icon

Share this with someone who'll enjoy it: