Provable Privacy Attacks on Trained Shallow Neural Networks
cs.LG, cs.CR
Submitted: 2024-10-10
Updated: 2026-08-26
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- One-shot Empirical Privacy Estimation for Federated Learning
- Information Complexity of Stochastic Convex Optimization: Applications to Generalization and Memorization
- Reconstructing Training Data from Multiclass Neural Networks
- Extracting Training Data from Diffusion Models
- Modelling and Quantifying Membership Information Leakage in Machine Learning
- An Extension of Fano's Inequality for Characterizing Model Susceptibility to Membership Inference Attacks
- Towards an Understanding of Benign Overfitting in Neural Networks
- Scalable Extraction of Training Data from (Production) Language Models
- Reconstructing Training Data From Real World Models Trained with Transfer Learning
- Investigating Trade-offs in Utility, Fairness and Differential Privacy in Neural Networks
- Diffusion Art or Digital Forgery? Investigating Data Replication in Diffusion Models
- Fishing for User Data in Large-Batch Federated Learning via Gradient Magnification
- Benign Overfitting and Grokking in ReLU Networks for XOR Cluster Data
- Leave-one-out Distinguishability in Machine Learning
- Understanding deep learning requires rethinking generalization
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: Local-Consistency Optimisation for Bayesian Neural Networks
- Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks