ICLR 2023 accepted papers on Federated learning and privacy

Personalized and heterogeneous FL · 20 papers

• Multimodal Federated Learning via Contrastive Representation Ensemble (36 citations)
• Personalized Federated Learning with Feature Alignment and Classifier Collaboration (34 citations)
• Better Generative Replay for Continual Federated Learning (13 citations)
• The Best of Both Worlds: Accurate Global and Personalized Models through Federated Learning with Data-Free Hyper-Knowledge Distillation (11 citations)
• Towards Understanding and Mitigating Dimensional Collapse in Heterogeneous Federated Learning (11 citations)
• On the Importance and Applicability of Pre-Training for Federated Learning (10 citations)
• FedFA: Federated Feature Augmentation (8 citations)
• FiT: Parameter Efficient Few-shot Transfer Learning for Personalized and Federated Image Classification (8 citations)
• Test-Time Robust Personalization for Federated Learning (7 citations)
• FedDAR: Federated Domain-Aware Representation Learning (4 citations)
• Federated Nearest Neighbor Machine Translation (3 citations)
• Combating Exacerbated Heterogeneity for Robust Models in Federated Learning (2 citations)

and 8 more in the interactive map.

Communication-efficient FL · 11 papers

• Where to Begin? On the Impact of Pre-Training and Initialization in Federated Learning (14 citations)
• Single-shot General Hyper-parameter Optimization for Federated Learning (13 citations)
• FedSpeed: Larger Local Interval, Less Communication Round, and Higher Generalization Accuracy (12 citations)
• FedExP: Speeding Up Federated Averaging via Extrapolation (7 citations)
• Sparse Random Networks for Communication-Efficient Federated Learning (7 citations)
• Meta Knowledge Condensation for Federated Learning (5 citations)
• Faster federated optimization under second-order similarity (2 citations)
• SWIFT: Rapid Decentralized Federated Learning via Wait-Free Model Communication (2 citations)
• Federated Learning as Variational Inference: A Scalable Expectation Propagation Approach (1 citation)
• Generalization Bounds for Federated Learning: Fast Rates, Unparticipating Clients and Unbounded Losses
• MocoSFL: enabling cross-client collaborative self-supervised learning

Differential privacy · 11 papers

• Disparate Impact in Differential Privacy from Gradient Misalignment (7 citations)
• Share Your Representation Only: Guaranteed Improvement of the Privacy-Utility Tradeoff in Federated Learning (6 citations)
• Exploring the Limits of Differentially Private Deep Learning with Group-wise Clipping (4 citations)
• Differentially Private Adaptive Optimization with Delayed Preconditioners (3 citations)
• Easy Differentially Private Linear Regression (3 citations)
• Individual Privacy Accounting with Gaussian Differential Privacy (2 citations)
• Private Federated Learning Without a Trusted Server: Optimal Algorithms for Convex Losses (2 citations)
• Statistical Theory of Differentially Private Marginal-based Data Synthesis Algorithms (1 citation)
• Stochastic Differentially Private and Fair Learning (1 citation)
• Differentially Private $L_2$-Heavy Hitters in the Sliding Window Model
• Regression with Label Differential Privacy

Privacy attacks and secure learning · 7 papers

• FLIP: A Provable Defense Framework for Backdoor Mitigation in Federated Learning (16 citations)
• Decepticons: Corrupted Transformers Breach Privacy in Federated Learning for Language Models (10 citations)
• Machine Unlearning of Federated Clusters (5 citations)
• Bias Propagation in Federated Learning (4 citations)
• CANIFE: Crafting Canaries for Empirical Privacy Measurement in Federated Learning (2 citations)
• Effective passive membership inference attacks in federated learning against overparameterized models
• Panning for Gold in Federated Learning: Targeted Text Extraction under Arbitrarily Large-Scale Aggregation