A comprehensive privacy assessment of LLMs.
A comprehensive trustworthiness assessment of compressed LLMs.
Zeroth-order optimization for LLM.
Recent advances in unsupervised learning have shown that unsupervised pre-training, followed by fine-tuning, can improve model generalization. However, a rigorous understanding of how the representation function learned on an unlabeled dataset …
A new method for safely and robustly injecting watermark after training without training data.
We propose a new risk to published generative models that finetuning on generated samples can exacerbate the privacy leakage.
We make local LLMs to engineer privacy-preserving prompts that are transferrable for cloud models.
Tracking IP leakage in federated learning.
We propose a new metric to efficiently evaluate the privacy risks from gradient inversion and provides new insights.
We uncover the security risk of data-free distillation from a poisoned teacher and propose the first countermeasure.