Downgrading optimizers to lower-information variants during LLM unlearning yields more robust forgetting on MUSE and WMDP benchmarks by converging to harder-to-perturb loss basins.
Proactive privacy amnesia for large language models: Safeguarding pii with negligible impact on model utility
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
years
2025 2representative citing papers
PeCL applies token-level dynamic differential privacy and privacy-guided memory sculpting to achieve superior privacy-utility balance in continual learning.
citing papers explorer
-
Downgrade to Upgrade: Optimizer Simplification Enhances Robustness in LLM Unlearning
Downgrading optimizers to lower-information variants during LLM unlearning yields more robust forgetting on MUSE and WMDP benchmarks by converging to harder-to-perturb loss basins.
-
Forget What's Sensitive, Remember What Matters: Token-Level Differential Privacy in Memory Sculpting for Continual Learning
PeCL applies token-level dynamic differential privacy and privacy-guided memory sculpting to achieve superior privacy-utility balance in continual learning.