Zinan Lin

The paper proposes DP-GCL, a novel differentially private contrastive learning framework that improves representation learning on sensitive data by bounding gradient dependency through localized group-level negative sampling.

DP-SAPF introduces a saliency-aware parameter fine-tuning method that selectively identifies the most critical parameters for LoRA training, significantly improving the utility and fidelity of differentially private image synthesis while reducing computational cost.

The paper introduces PE-means, an improved differentially private $k$-means clustering method that uses the Private Evolution (PE) algorithm to achieve better clustering loss compared to existing state-of-the-art techniques.