machine unlearning

Causal Routing for Unlearning
LLMs cannot forget the way we delete a file. Strangely, we are asked to remove something that was never put anywhere in particular. …
Bardh Prenkaj, Andrea D'Angelo, Davide Mottin, Federico Fontana, Davide Gabrielli, Paola Velardi, Stefano Faralli
PDF Cite DOI
Beyond Binary Rewards: A Comparative Study of Reward Design for Reinforcement Unlearning
Machine unlearning seeks to selectively remove specific knowledge from trained language models without full retraining, a growing …
Efstratios Zaradoukas, Davide Gabrielli, Bardh Prenkaj, Gjergji Kasneci
PDF Cite DOI
Reinforcement Unlearning via Group Relative Policy Optimization
During pretraining, LLMs inadvertently memorize sensitive or copyrighted data, posing significant compliance challenges under legal …
Efstratios Zaradoukas, Bardh Prenkaj, Gjergji Kasneci
PDF Cite DOI
CURE: Controlled Unlearning for Robust Embeddings - Mitigating Conceptual Shortcuts in Pre-Trained Language Models
Pre-trained language models have achieved remarkable success across diverse applications but remain susceptible to spurious, …
Aysenur Kocak, Shuo Yang, Bardh Prenkaj, Gjergji Kasneci
PDF Cite DOI