安全聚合论文趋势 2023–2026

主要结论

• 安全聚合论文从 2023 年的 14 篇增长到 2026 年的 25 篇，三年间扩大到约 1.8 倍。
• NeurIPS 是发安全聚合论文最多的会议，2023–2026 年共 27 篇，占四个会议总量的 35%。
• 交叉方向里涨得最快的是驾驶仿真：同时属于两个主题的论文从 2023 年的 3 篇增至 2026 年的 5 篇。
• 在「联邦与隐私」的 5 个主题里，它的 2026 年论文数排第 5；同组增长最快的隐私攻击是 1.8 倍，安全聚合是 1.8 倍。

代表论文

按年份和会议均匀挑选，方便快速看到每一届的样貌

• NeurIPS 2026Can Model Merging Improve Aggregation in DiLoCo?
• ICML 2026BESplit: Bias-Compensated Split Federated Learning with Evidential Aggregation
• ICLR 2026Beyond Aggregation: Guiding Clients in Heterogeneous Federated Learning
• CVPR 2026A2GC Asymmetric Aggregation with Geometric Constraints for Locally Aggregated Descriptors
• NeurIPS 2025Revolutionizing Graph Aggregation: From Suppression to Amplification via BoostGCN
• ICML 2025Janus: Dual-Server Multi-Round Secure Aggregation with Verifiability for Federated Learning
• ICLR 2025Divergence-Regularized Discounted Aggregation: Equilibrium Finding in Multiplayer Partially Observable Stochastic Games
• CVPR 2025FedCALM: Conflict-aware Layer-wise Mitigation for Selective Aggregation in Deeper Personalized Federated Learning
• NeurIPS 2024RFLPA: A Robust Federated Learning Framework against Poisoning Attacks with Secure Aggregation
• ICML 2024Converting Transformers to Polynomial Form for Secure Inference Over Homomorphic Encryption
• CVPR 2024GoMVS: Geometrically Consistent Cost Aggregation for Multi-View Stereo
• NeurIPS 2023Secure Out-of-Distribution Task Generalization with Energy-Based Models
• ICML 2023Secure Federated Correlation Test and Entropy Estimation
• ICLR 2023Cycle to Clique (Cy2C) Graph Neural Network: A Sight to See beyond Neighborhood Aggregation
• CVPR 2023SliceMatch: Geometry-Guided Aggregation for Cross-View Pose Estimation
• NeurIPS 2026High-Dimensional Conditional Independence Testing via Random Projection Aggregation
• ICML 2026Let the Prototype Guide You: Robust Aggregation of Sparse Multi-Class Annotations via Annotator Prototype Learning
• ICLR 2026Decision Aggregation under Quantal Response
• CVPR 2026Factorized Context Aggregation for Robust Cancer Risk Estimation via Soft
• NeurIPS 2025Sound Logical Explanations for Mean Aggregation Graph Neural Networks

全部安全聚合论文

共 20 篇，第 1 / 1 页

1. ICML 2026BESplit: Bias-Compensated Split Federated Learning with Evidential Aggregation
2. ICML 2026DAG-MoE: From Simple Mixture to Structural Aggregation in Mixture-of-Experts
3. ICML 2026Fixed Aggregation Features Can Rival GNNs
4. ICML 2026Let the Prototype Guide You: Robust Aggregation of Sparse Multi-Class Annotations via Annotator Prototype Learning
5. ICML 2026Set-Preserving Calibration from Conformal P-Values to E-Values
6. ICML 2025FicGCN: Unveiling the Homomorphic Encryption Efficiency from Irregular Graph Convolutional Networks
7. ICML 2025Janus: Dual-Server Multi-Round Secure Aggregation with Verifiability for Federated Learning
8. ICML 2025SecEmb: Sparsity-Aware Secure Federated Learning of On-Device Recommender System with Large Embedding
9. ICML 2024Converting Transformers to Polynomial Form for Secure Inference Over Homomorphic Encryption
10. ICML 2024Ditto: Quantization-aware Secure Inference of Transformers upon MPC
11. ICML 2024Provably Better Explanations with Optimized Aggregation of Feature Attributions
12. ICML 2024Resisting Stochastic Risks in Diffusion Planners with the Trajectory Aggregation Tree
13. ICML 2024Revisiting Context Aggregation for Image Matting
14. ICML 2024Self-Driven Entropy Aggregation for Byzantine-Robust Heterogeneous Federated Learning
15. ICML 2024Structure-Aware E(3)-Invariant Molecular Conformer Aggregation Networks
16. ICML 2023Flexible Model Aggregation for Quantile Regression
17. ICML 2023PINA: Leveraging Side Information in eXtreme Multi-label Classification via Predicted Instance Neighborhood Aggregation
18. ICML 2023Secure Federated Correlation Test and Entropy Estimation
19. ICML 2023SegCLIP: Patch Aggregation with Learnable Centers for Open-Vocabulary Semantic Segmentation
20. ICML 2023SRATTA: Sample Re-ATTribution Attack of Secure Aggregation in Federated Learning.

常见问题

2023–2026 年四大会议一共有多少篇安全聚合论文？

共 78 篇，按年份分别是 2023 年 14 篇、2024 年 20 篇、2025 年 19 篇、2026 年 25 篇。

NeurIPS 2026 有多少篇安全聚合论文？

NeurIPS 2026 有 8 篇安全聚合论文，上一届是 11 篇。

ICLR 2026 有多少篇安全聚合论文？

ICLR 2026 有 7 篇安全聚合论文，上一届是 1 篇。

ICML 2026 有多少篇安全聚合论文？

ICML 2026 有 5 篇安全聚合论文，上一届是 3 篇。

CVPR 2026 有多少篇安全聚合论文？

CVPR 2026 有 5 篇安全聚合论文，上一届是 4 篇。

哪个会议的安全聚合论文最多？

NeurIPS，2023–2026 年共 27 篇。