Unlearning Reveals the Influential Training Data of Language Models

Open in new window