ICLR 2026 accepted papers on Federated learning and privacy
81 of the 5,358 papers accepted at ICLR 2026 (1.5%) are on Federated learning and privacy, compared with 1.7% at ICLR 2025. They fall into 4 topics.
See ICLR 2026 in the interactive map
Working on something in this area? Describe your idea in scime atlas to see the closest accepted papers by meaning and how crowded the spot has become.
Topics and papers
Differential privacy · 31 papers
- Back to Square Roots: An Optimal Bound on the Matrix Factorization Error for Multi-Epoch Differentially Private SGD (1 citation)
- Adaptive Methods Are Preferable in High Privacy Settings: An SDE Perspective
- Benchmarking Empirical Privacy Protection for Adaptations of Large Language Models
- Beyond Membership: Limitations of Add/Remove Adjacency in Differential Privacy
- Convergent Differential Privacy Analysis for General Federated Learning
- Differentially Private Equilibrium Finding in Polymatrix Games
- Differentially Private Two-Stage Gradient Descent for Instrumental Variable Regression
- Gaussian certified unlearning in high dimensions: A hypothesis testing approach
- INO-SGD: Addressing Utility Imbalance under Individualized Differential Privacy
- Natural Identifiers for Privacy and Data Audits in Large Language Models
- On Optimal Hyperparameters for Differentially Private Deep Transfer Learning
- Optimizing Canaries for Privacy Auditing with Metagradient Descent
and 19 more in the interactive map.
Personalized and heterogeneous FL · 29 papers
- Towards Privacy-Guaranteed Label Unlearning in Vertical Federated Learning: Few-Shot Forgetting Without Disclosure (1 citation)
- A Federated Generalized Expectation-Maximization Algorithm for Mixture Models with an Unknown Number of Components
- Deep Latent Variable Model based Vertical Federated Learning with Flexible Alignment and Labeling Scenarios
- Deploying Models to Non-participating Clients in Federated Learning without Fine-tuning: A Hypernetwork-based Approach
- FedDAG: Clustered Federated Learning via Global Data and Gradient Integration for Heterogeneous Environments
- Federated ADMM from Bayesian Duality
- Federated Learning with Profile Mapping under Distribution Shifts and Drifts
- FLoRG: Federated Fine-tuning with Low-rank Gram Matrices and Procrustes Alignment
- Heterogeneous Federated Fine-Tuning with Parallel One-Rank Adaptation
- Low-pass Personalized Subgraph Federated Recommendation
- Mitigating Non-IID Drift in Zeroth-Order Federated LLM Fine-Tuning with Transferable Sparsity
- Not All Clients Are Equal: Collaborative Model Personalization on Heterogeneous Multi-Modal Clients
and 17 more in the interactive map.
Privacy attacks and secure learning · 11 papers
- Byzantine-Robust Federated Learning with Learnable Aggregation Weights
- Concept-Aware Privacy Mechanisms for Defending Embedding Inversion Attacks
- Curation Leaks: Membership Inference Attacks against Data Curation for Machine Learning
- Learnability and Privacy Vulnerability are Entangled in a Few Critical Weights
- Mitigating Privacy Risk via Forget Set-Free Unlearning
- Protection against Source Inference Attacks in Federated Learning
- Robust Federated Inference
- SABRE-FL: Selective and Accurate Backdoor Rejection for Federated Prompt Learning
- Traceable Black-Box Watermarks For Federated Learning
- Near Optimal Robust Federated Learning Against Data Poisoning Attack
- ULD-Net: Enabling Ultra-Low-Degree Fully Polynomial Networks for Homomorphically Encrypted Inference
Communication-efficient FL · 10 papers
- Converge Faster, Talk Less: Hessian-Informed Federated Zeroth-Order Optimization
- DeepAFL: Deep Analytic Federated Learning
- FedMuon: Federated Learning with Bias-corrected LMO-based Optimization
- Layerwise Federated Learning for Heterogeneous Quantum Clients using Quorus
- Non-Convex Federated Optimization under Cost-Aware Client Selection
- Riemannian Federated Learning via Averaging Gradient Streams
- Single-Loop Byzantine-Resilient Federated Bilevel Optimization
- Enhancing Communication Compression via Discrepancy-aware Calibration for Federated Learning
- Unified Analyses for Hierarchical Federated Learning: Topology Selection under Data Heterogeneity
- Unlocking the Potential of Weighting Methods in Federated Learning Through Communication Compression