Unlearned but Not Forgotten: Data Extraction after Exact Unlearning in LLM
Published in NeurIPS 2025, 2025
We revisit exact unlearning under realistic deployments that expose both pre- and post-unlearning models. Guidance from the original model and token filtering can recover information that the retrained model was intended to forget.
Recommended citation: Wu, X., Pang, Y., Liu, T., and Wu, Z. S. (2025). "Unlearned but Not Forgotten: Data Extraction after Exact Unlearning in LLM." NeurIPS 2025.
Download Paper