Less is More: Geometric Unlearning for LLMs with Minimal Data Disclosure

ArXi:2605.01735v1 Announce Type: new As large language models (LLMs) are increasingly deployed in real-world systems, they must post-hoc removal of specific content to meet privacy and governance requirements. This motivates selective unlearning, which suppresses information about a particular entity or topic while preserving the LLM's general utility. However, most existing LLM unlearning methods require access to the original