Hanlin Gu

Towards Privacy-Guaranteed Label Unlearning in Vertical Federated Learning: Few-Shot Forgetting Without Disclosure

FedSycle: Mitigating Post-Unlearning Performance Inconsistency in Federated Learning via Latent Feature Decoupling

LLEOT: A Privacy-Enhancing Offsite Tuning Framework via Loss Landscape Elevation

Order-Level Attention Similarity Across Language Models: A Latent Commonality

A few-shot Label Unlearning in Vertical Federated Learning

Disentangling data distribution for Federated Learning