分布式训练论文趋势 2023–2026

主要结论

• 分布式训练论文从 2023 年的 95 篇增长到 2026 年的 98 篇，三年间扩大到约 1.0 倍。
• NeurIPS 是发分布式训练论文最多的会议，2023–2026 年共 208 篇，占四个会议总量的 53%。
• 交叉方向里涨得最快的是多模态大模型：同时属于两个主题的论文从 2023 年的 1 篇增至 2026 年的 10 篇。
• 在「效率与系统」的 5 个主题里，它的 2026 年论文数排第 4；同组增长最快的硬件协同是 4.4 倍，分布式训练是 1.0 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 2026AgentWeave: Efficient Distributed Agent Serving via Flow Decomposition
• ICML 2026Breaking the Capacity Bottleneck in Model-Heterogeneous Federated Learning via Gradual Model Restoration
• ICLR 2026Cascadia: An Efficient Cascade Serving System for Large Language Models
• CVPR 2026Finding Distributed Object-Centric Properties in Self-Supervised Transformers
• NeurIPS 2025A Fair Federated Learning Method for Handling Client Participation Probability Inconsistencies in Heterogeneous Environments
• ICML 2025BSemiFL: Semi-supervised Federated Learning via a Bayesian Approach
• ICLR 2025Addressing Label Shift in Distributed Learning via Entropy Regularization
• CVPR 2025Real-time High-fidelity Gaussian Human Avatars with Position-based Interpolation of Spatially Distributed MLPs
• NeurIPS 2024Communication Efficient Distributed Training with Distributed Lion
• ICML 2024Embarrassingly Parallel GFlowNets
• ICLR 2024Benchmarking Algorithms for Federated Domain Generalization
• CVPR 2024Deciphering 'What' and 'Where' Visual Pathways from Spectral Clustering of Layer-Distributed Neural Representations
• NeurIPS 20232Direction: Theoretically Faster Distributed Training with Bidirectional Communication Compression
• ICML 2023Block Subsampled Randomized Hadamard Transform for Nyström Approximation on Distributed Architectures
• ICLR 2023Distributed Extra-gradient with Optimal Complexity and Communication Guarantees
• CVPR 2023DisCo-CLIP: A Distributed Contrastive Loss for Memory Efficient CLIP Training
• NeurIPS 2026Do We Need Asynchronous SGD? On the Near-Optimality of Synchronous Methods
• ICML 2026Clipping Makes Distributed and Federated Asynchronous SGD Robust to Stragglers
• ICLR 2026DualMap: Enabling Both Cache Affinity and Load Balancing for Distributed LLM Serving
• CVPR 2026Parallel Rigidity Matters for Bundle Adjustment

全部分布式训练论文

共 42 篇，第 1 / 1 页

1. ICML 2026A Short and Unified Convergence Analysis of the SAG, SAGA, and IAG Algorithms
2. ICML 2026A Tight Theory of Error Feedback Algorithms in Distributed Optimization
3. ICML 2026Accelerated Dual Method for Distributed Optimization: An Inexact-Gradient View of Local Updates
4. ICML 2026Adaptive Batch Sizes Using Non-Euclidean Gradient Noise Scales for Stochastic Sign and Spectral Descent
5. ICML 2026Breaking the Capacity Bottleneck in Model-Heterogeneous Federated Learning via Gradual Model Restoration
6. ICML 2026Clipping Makes Distributed and Federated Asynchronous SGD Robust to Stragglers
7. ICML 2026CoCoQuant: Breaking the Bandwidth Wall via Co-Optimized Communication and Computation Quantization
8. ICML 2026CONCUR: High-Throughput Agentic Batch Inference of LLM via Congestion-Based Concurrency Control
9. ICML 2026Convergence Analysis of the Lion Optimizer in Centralized and Distributed Settings
10. ICML 2026Decoupled Low-Rank Adaptation for Robust Federated Fine-Tuning
11. ICML 2026Delayed Momentum Aggregation: Communication-efficient Byzantine-robust Federated Learning with Partial Participation
12. ICML 2026DistFlow: A Fully Distributed RL Framework for Scalable and Efficient LLM Post-Training
13. ICML 2026Distributed Stochastic $K$-Level Optimization Over Networks
14. ICML 2026DITRON: Distributed Multi-level Tiling Compiler for Parallel Tensor Programs
15. ICML 2026Dynamics of neural scaling laws in random feature regression with powerlaw-distributed kernel eigenvalues
16. ICML 2026EasyBalance: Cross-Layer Load Balancing in Distributed MoE Inference
17. ICML 2026Efficient Distributed MLLM Training with Cornstarch
18. ICML 2026FAB: A First-Order AB-based Gradient Algorithm for Distributed Bilevel Optimization over Time-Varying Directed Graphs
19. ICML 2026Factored Gossip DiLoCo: Reducing Blocking Communication within DiLoCo
20. ICML 2026FedARC: Anchor-Guided Residual Compensation for Data and Model Heterogeneous Federated Learning
21. ICML 2026FedCDWA: Decoupled Federated Prototype Distillation with Hierarchical Wasserstein Aggregation
22. ICML 2026FedSSM: State Space Model-based Proactive Inference for Heterogeneous Multimodal Federated Learning
23. ICML 2026Fisher-Preserving Guidance: Training-Free Manifold Constraints for Safe Diffusion Control
24. ICML 2026gp2Scale: A Class of Compactly Supported Non-Stationary Kernels and Distributed Computing for Exact Gaussian Processes on 10 Million Data Points
25. ICML 2026HexGen-3: A Fully Disaggregated LLM Serving Framework with Fine-Grained Heterogeneous Resource Autoscaling
26. ICML 2026High-Probability Convergence Guarantees of Decentralized SGD
27. ICML 2026Identifying and Mitigating Errors in Gradient Aggregation of Distributed Data Parallel Training
28. ICML 2026JanusPipe: Efficient Pipeline Parallel Training for Machine Learning Interatomic Potentials
29. ICML 2026Least-Loaded Expert Parallelism: Load Balancing An Imbalanced Mixture-of-Experts
30. ICML 2026LoRDO: Distributed Low-Rank Optimization with Infrequent Communication
31. ICML 2026Matrix-Free GPU Semidefinite Programming for Quantum Ordered Search at the k=6 Frontier
32. ICML 2026Mitigating Staleness in Asynchronous Pipeline Parallelism via Basis Rotation
33. ICML 2026MuLoCo: Muon is a Practical Inner Optimizer for DiLoCo
34. ICML 2026On the Role of Batch Size in Stochastic Conditional Gradient Methods
35. ICML 2026One-Step Gradient Delay is Not a Barrier for Large-Scale Asynchronous Pipeline Parallel LLM Pretraining
36. ICML 2026OServe: Accelerating LLM Serving via Spatial-Temporal Workload Orchestration
37. ICML 2026Random Scaling of Emergent Capabilities
38. ICML 2026Robust Federated Learning Against Adaptive Compression
39. ICML 2026Row-Stochastic Matrices Can Provably Outperform Doubly Stochastic Matrices in Decentralized Learning
40. ICML 2026scDataset: Scalable Data Loading for Deep Learning on Large-Scale Single-Cell Omics
41. ICML 2026SFedPO: Streaming Federated Learning with a Prediction Oracle under Temporal Shifts
42. ICML 2026Tight Stability Bounds for Robust Distributed Learning: Byzantine Failures Hurt Generalization More than Data Poisoning

常见问题

2023–2026 年四大会议一共有多少篇分布式训练论文？

共 396 篇，按年份分别是 2023 年 95 篇、2024 年 74 篇、2025 年 129 篇、2026 年 98 篇。

NeurIPS 2026 有多少篇分布式训练论文？

NeurIPS 2026 有 36 篇分布式训练论文，上一届是 69 篇。

ICLR 2026 有多少篇分布式训练论文？

ICLR 2026 有 18 篇分布式训练论文，上一届是 18 篇。

ICML 2026 有多少篇分布式训练论文？

ICML 2026 有 42 篇分布式训练论文，上一届是 40 篇。

CVPR 2026 有多少篇分布式训练论文？

CVPR 2026 有 2 篇分布式训练论文，上一届是 2 篇。

哪个会议的分布式训练论文最多？

NeurIPS，2023–2026 年共 208 篇。