1 paper
Noam Diamant, Neta Glazer, Ethan Fetaya
Recently, machine unlearning, the removal of specific training data influence from a model, has gained increasing attention. In large language models (LLMs), unlearning is particul…