mlsanityData to 14 Oct 2026

VP-NTK: Exploring the Benefits of Visual Prompting in Differentially Private Data Synthesis

arXiv 2503.16195 ↗First posted 20 Mar 2025cs.CV cs.LGcited by 0

Not on the map. Why?

Abstract (arXiv)

Differentially private (DP) synthetic data has become the de facto standard for releasing sensitive data. However, many DP generative models suffer from the low utility of synthetic data, especially for high-resolution images. On the other hand, one of the emerging techniques in parameter efficient fine-tuning (PEFT) is visual prompting (VP), which allows well-trained existing models to be reused for the purpose of adapting to subsequent downstream tasks. In this work, we explore such a phenomenon in constructing captivating generative models with DP constraints. We show that VP in conjunction with DP-NTK, a DP generator that exploits the power of the neural tangent kernel (NTK) in training DP generative models, achieves a significant performance boost, particularly for high-resolution image datasets, with accuracy improving from 0.644±0.044 to 0.769. Lastly, we perform ablation studies on the effect of different parameters that influence the overall performance of VP-NTK. Our work demonstrates a promising step forward in improving the utility of DP synthetic data, particularly for high-resolution images.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 18 of its 22 references that are in the map. First posted on arXiv 20 Mar 2025.

Builds on

18 of 22 references.

References · 18

  1. 2021Pre-train, Prompt, and Predict: A Systematic Survey of Prompting Methods in Natural Language Processingcited by 2,234
  2. 2022Exploring Visual Prompts for Adapting Large-Scale Modelscited by 274
  3. 2019Differentially Private Learning with Adaptive Clippingcited by 220
  4. 2021Instance-Conditioned GANcited by 121
  5. 2022Differentially Private Diffusion Modelscited by 119
  6. 2022Understanding and Improving Visual Prompting: A Label-Mapping Perspectivecited by 87
  7. 2021Do Not Let Privacy Overbill Utility: Gradient Embedding Perturbation for Private Learningcited by 86
  8. 2018A General Approach to Adding Differential Privacy to Iterative Training Procedurescited by 84
  9. 2020Transfer Learning without Knowing: Reprogramming Black-box Machine Learning Models with Scarce Data and Limited Resourcescited by 83
  10. 2020DP-MERF: Differentially Private Mean Embeddings with Random Features for Practical Privacy-Preserving Data Generationcited by 78
  11. 2021Don't Generate Me: Training Differentially Private Generative Models with Sinkhorn Divergencecited by 54
  12. 2019G-PATE: Scalable Differentially Private Data Generator via Private Aggregation of Teacher Discriminatorscited by 53
6 more
  1. 2021Cross-modal Adversarial Reprogrammingcited by 34
  2. 2021DataLens: Scalable Privacy Preserving Training via Gradient Compression and Aggregationcited by 31
  3. 2021PEARL: Data Synthesis via Private Embeddings and Adversarial Reconstruction Learningcited by 23
  4. 2021Hermite Polynomial Features for Private Data Generationcited by 21
  5. 2023Exploring the Benefits of Visual Prompting in Differential Privacycited by 17
  6. 2023Differentially Private Neural Tangent Kernels for Privacy-Preserving Data Generationcited by 15

Built on by

Not cited yet.