Output forgetting in machine unlearning overestimates success because unlearned models exhibit structured representation mismatches relative to retraining from scratch.
On the impossibility of retrain equivalence in machine unlearning.arXiv preprint arXiv:2510.16629, 2025
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Erased, but Not Gone: Output Forgetting Is Not True Forgetting
Output forgetting in machine unlearning overestimates success because unlearned models exhibit structured representation mismatches relative to retraining from scratch.