Behavioral Guarantees for Proxy-Based Unlearning

Excerpt

arXiv:2605.10680v2 Announce Type: replace Abstract: This paper proposes a framework generalizing recent proxy-based unlearning methods and proves theoretical guarantees about the behavior of the resulting unlearned model: upper bounds on its Kullback-Leibler divergence to the ideal posterior distribution of the retain data. We model approximate unlearning as a constrained optimization problem and interpret a family of solutions as introducing a scaled unlearning signal in the output space. The u