NeuroImprint attack assigns isolated memorization neurons to training samples in PEFT adapters, enabling closed-form reconstruction of 59-79% of samples across BERT, GPT-2, Qwen2, and Llama3.2 on multiple datasets.
Fishing for user data in large-batch federated learning via gra- dient magnification
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.CR 2years
2026 2representative citing papers
FLRSP enhances privacy in federated learning by randomly selecting model parameters for sharing, delivering competitive image classification accuracy and improved resistance to reconstruction attacks on ResNet34 and ViT models using FedSGD and FedAvg.
citing papers explorer
-
From Efficiency to Leakage -- Privacy Backdoor in Federated Language Model Fine-Tuning
NeuroImprint attack assigns isolated memorization neurons to training samples in PEFT adapters, enabling closed-form reconstruction of 59-79% of samples across BERT, GPT-2, Qwen2, and Llama3.2 on multiple datasets.
-
FLRSP: Privacy-Preserving Federated Learning Using Randomly Selected Model Parameters
FLRSP enhances privacy in federated learning by randomly selecting model parameters for sharing, delivering competitive image classification accuracy and improved resistance to reconstruction attacks on ResNet34 and ViT models using FedSGD and FedAvg.