NeurIPS 2022 accepted papers on Federated learning and privacy
86 of the 2,901 papers accepted at NeurIPS 2022 (3.0%) are on Federated learning and privacy, compared with 2.9% at NeurIPS 2021. They fall into 4 topics.
See NeurIPS 2022 in the interactive map
Submitting to NeurIPS? Checklist →
Working on something in this area? Describe your idea in scime atlas to see the closest papers by meaning and how crowded the spot has become.
Topics and papers
Differential privacy · 37 papers
- On Privacy and Personalization in Cross-Silo Federated Learning (20 citations)
- Private Set Generation with Discriminative Information (13 citations)
- Differentially Private Linear Sketches: Efficient Implementations and Applications (11 citations)
- Improved Differential Privacy for SGD via Optimal Private Linear Operators on Adaptive Streams (8 citations)
- Scalable and Efficient Training of Large Convolutional Neural Networks with Differential Privacy (8 citations)
- When Does Differentially Private Learning Not Suffer in High Dimensions? (8 citations)
- Composition Theorems for Interactive Differential Privacy (6 citations)
- New Lower Bounds for Private Estimation and a Generalized Fingerprinting Lemma (6 citations)
- DP-PCA: Statistically Optimal and Differentially Private PCA (5 citations)
- Order-Invariant Cardinality Estimators Are Differentially Private (5 citations)
- Privacy Induces Robustness: Information-Computation Gaps and Sparse Mean Estimation (5 citations)
- Differentially Private Learning Needs Hidden State (Or Much Faster Convergence) (4 citations)
and 25 more in the interactive map.
Personalized and heterogeneous FL · 20 papers
- Federated Learning from Pre-Trained Models: A Contrastive Learning Approach (65 citations)
- FedRolex: Model-Heterogeneous Federated Learning with Rolling Sub-Model Extraction (45 citations)
- Preservation of the Global Knowledge by Not-True Distillation in Federated Learning (42 citations)
- SemiFL: Semi-Supervised Federated Learning for Unlabeled Clients with Alternate Training (36 citations)
- DENSE: Data-Free One-Shot Federated Learning (33 citations)
- pFL-Bench: A Comprehensive Benchmark for Personalized Federated Learning (23 citations)
- DReS-FL: Dropout-Resilient Secure Federated Learning for Non-IID Clients via Secret Data Sharing (19 citations)
- FairVFL: A Fair Vertical Federated Learning Framework with Contrastive Adversarial Learning (17 citations)
- CalFAT: Calibrated Federated Adversarial Training with Label Skewness (14 citations)
- FedSR: A Simple and Effective Domain Generalization Method for Federated Learning (13 citations)
- An Adaptive Kernel Approach to Federated Learning of Heterogeneous Causal Effects (7 citations)
- A Coupled Design of Exploiting Record Similarity for Practical Vertical Federated Learning (5 citations)
and 8 more in the interactive map.
Communication-efficient FL · 16 papers
- FedAvg with Fine Tuning: Local Updates Lead to Representation Learning (46 citations)
- A Unified Analysis of Federated Learning with Arbitrary Client Participation (16 citations)
- Fairness in Federated Learning via Core-Stability (11 citations)
- SoteriaFL: A Unified Framework for Private Federated Learning with Communication Compression (10 citations)
- TCT: Convexifying Federated Learning using Bootstrapped Neural Tangent Kernels (5 citations)
- FedPop: A Bayesian Approach for Personalised Federated Learning (3 citations)
- Personalized Federated Learning towards Communication Efficiency, Robustness and Fairness (3 citations)
- Resource-Adaptive Federated Learning with All-In-One Neural Composition (3 citations)
- Taming Fat-Tailed (“Heavier-Tailed” with Potentially Infinite Variance) Noise in Federated Learning (3 citations)
- Outsourcing Training without Uploading Data via Efficient Collaborative Open-Source Sampling (2 citations)
- Personalized Online Federated Learning with Multiple Kernels (2 citations)
- Variance Reduced ProxSkip: Algorithm, Theory and Application to Federated Learning (2 citations)
and 4 more in the interactive map.
Privacy attacks and secure learning · 13 papers
- FLamby: Datasets and Benchmarks for Cross-Silo Federated Learning in Realistic Healthcare Settings (73 citations)
- LAMP: Extracting Text from Gradients with Language Model Priors (21 citations)
- Recovering Private Text in Federated Learning of Language Models (21 citations)
- Iron: Private Inference on Transformers (18 citations)
- The Privacy Onion Effect: Memorization is Relative (18 citations)
- Learning to Attack Federated Learning: A Model-based Reinforcement Learning Attack Framework (13 citations)
- A General Framework for Auditing Differentially Private Machine Learning (6 citations)
- ConfLab: A Data Collection Concept, Dataset, and Benchmark for Machine Analysis of Free-Standing Social Interactions in the Wild (4 citations)
- SecureFedYJ: a safe feature Gaussianization protocol for Federated Learning (3 citations)
- Byzantine-tolerant federated Gaussian process regression for streaming data (1 citation)
- Homomorphic Matrix Completion
- Private Multiparty Perception for Navigation
and 1 more in the interactive map.
