Knowledge Injection And Unlearning

Papers on selectively inserting, editing, or removing specific knowledge from language models while preserving existing capabilities and generalization.