Split, Unlearn, Merge: Leveraging Data Attributes for More Effective Unlearning in LLMs

Devs

Split, Unlearn, Merge: Leveraging Data Attributes for More Effective Unlearning in LLMs | Read Paper on Bytez