TY - RPRT TI - UNDIAL: Self-Distillation with Adjusted Logits for Robust Unlearning in Large Language Models AU - Yijiang River Dong AU - Hongzhou Lin AU - Mikhail Belkin AU - Ramon Huerta AU - Ivan Vulić PY - 2024 UR - https://arxiv.org/abs/2402.10052 ID - 2402.10052 ER -