Who's Harry Potter? Approximate Unlearning in LLMs
Paper • 2310.02238 • Published
Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
Genericized Harry Potter tokens for the Who's Harry Potter Unlearning method, from https://arxiv.org/pdf/2310.02238. NOT MY NOVEL METHOD