Redirection for Erasing Memory (REM): Towards a universal unlearning method for corrupted data

ICLR 2026 Conference Submission12426 Authors

Keywords: machine unlearning

TL;DR: We present a taxonomy for corrupted data unlearning tasks along two dimensions: regularity and discovery rate, and show that no prior method succeeds in all but a slice of that space. Our method is the first to perform strongly across this space.

Abstract: Machine unlearning is studied for a multitude of tasks, but specialization of unlearning methods to particular tasks has made their systematic comparison challenging. To address this issue, we propose a conceptual space to characterize diverse corrupted data unlearning tasks in vision classifiers. This space is described by two dimensions, the discovery rate (the fraction of the corrupted data that are known at unlearning time) and the statistical regularity of the corrupted data (from random exemplars to shared concepts). Methods proposed previously have been targeted at portions of this space and, as we show, fail predictably outside these regions. We propose Redirection for Erasing Memory (REM), whose key feature is that corrupted data are redirected to dedicated neurons introduced at unlearning time and then discarded or deactivated to suppress the influence of corrupted data. REM performs strongly across the space of tasks, in contrast to prior SOTA methods that fail outside the regions for which they were designed.

Primary Area: other topics in machine learning (i.e., none of the above)

Submission Number: 12426