Second-order optimizers retain residual geometric memory in their state after unlearning that first-order metrics miss, and only controlled eigendecay perturbations fully erase it.
When machine unlearning jeopardizes privacy
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.LG 2years
2026 2verdicts
UNVERDICTED 2representative citing papers
Machine unlearning for online L-BFGS requires aligning the full optimizer state including memory to a counterfactual history without deleted samples rather than parameter correction alone.
citing papers explorer
-
Shape of Memory: a Geometric Analysis of Machine Unlearning in Second-Order Optimizers
Second-order optimizers retain residual geometric memory in their state after unlearning that first-order metrics miss, and only controlled eigendecay perturbations fully erase it.
-
Form and Function: Machine Unlearning as a Problem of Misaligned States
Machine unlearning for online L-BFGS requires aligning the full optimizer state including memory to a counterfactual history without deleted samples rather than parameter correction alone.