Reload performs machine unlearning without access to the forget set by combining gradient optimization with structured weight sparsification, approximating retrained models while using minimal retain-set data and model weights.
Title resolution pending
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
-
Mitigating Privacy Risk via Forget Set-Free Unlearning
Reload performs machine unlearning without access to the forget set by combining gradient optimization with structured weight sparsification, approximating retrained models while using minimal retain-set data and model weights.