References: OpenUnlearning(https://github.com/locuslab/open-unlearning), TOFU (https://github.com/locuslab/tofu), MUSE (https://github.com/swj0419/muse_bench), WMDP (https://github.com/centerforaisafety/wmdp), LMEvalHarness (https://github.com/EleutherAI/lm-evaluation-harness)
