Maliciously crafted deletion requests can make an unlearned model overconfident or underconfident on target samples while preserving its labels, across multiple uncertainty quantification methods.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Towards Unveiling Predictive Uncertainty Vulnerabilities in the Context of the Right to Be Forgotten
Maliciously crafted deletion requests can make an unlearned model overconfident or underconfident on target samples while preserving its labels, across multiple uncertainty quantification methods.