Understanding Private Evolution as Learning-Augmented Clustering
cs.LG, cs.CR, stat.ML
Submitted: 2026-09-29
Updated: 2026-09-29
Comments: to be published in NeurIPS 2026
Code: https://github.com/microsoft/DPSDA
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- Differentially Private Diffusion Models Generate Useful Synthetic Images
- Private Evolution Converges
- Feature Likelihood Divergence: Evaluating the Generalization of Generative Models Using Samples
- Harnessing large-language models to generate private synthetic text
- Differentially Private Synthetic Data via APIs 3: Using Simulators Instead of Foundation Model
- Differentially Private Synthetic Data: Applied Evaluations and Enhancements
- Is API Access to LLMs Useful for Generating Private Synthetic Tabular Data?
- Minimax Distribution Estimation in Wasserstein Distance
- Synthesize Privacy-Preserving High-Resolution Images via Private Textual Intermediaries
- Struct-Bench: A Benchmark for Differentially Private Structured Text Generation
- Private Training & Data Generation by Clustering Embeddings
Related papers
- Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
- AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
- Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
- Convergence issues in Relational Concept Analysis based on AOC-posets
- Beliefs Beyond Posteriors: The Predictive Cost of ELBO Binding
- Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks