GRAIL: Gradient-Based Adaptive Unlearning for Privacy and Copyright in LLMs.

Kun-Woo Kim, Ji-Hoon Park, Ju-Min Han, Seong-Whan Lee

Browse the full IJCNN paper archive.