分布式训练论文趋势 2023–2026

并行与大规模训练

统计范围：NeurIPS、ICLR、ICML、CVPR 四个会议，2023–2026 年，共 396 篇被归入该主题的论文。

主要结论

每年论文数

各会议论文数

按年份和会议统计

分布式训练论文数，按年份和会议
年份NeurIPSICLRICMLCVPR合计
2023541028395
202449128574
20256918402129
2026361842298

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

全部分布式训练论文

年份全部2023202420252026
会议全部NeurIPSICLRICMLCVPR

共 396 篇，第 4 / 8 页

  1. ICML 2025Ringmaster ASGD: The First Asynchronous SGD with Optimal Time Complexity
  2. ICML 2025Safe-EF: Error Feedback for Non-smooth Constrained Optimization
  3. ICML 2025Scalable First-order Method for Certifying Optimal k-Sparse GLMs
  4. ICML 2025Shortcut-connected Expert Parallelism for Accelerating Mixture of Experts
  5. ICML 2025SPD: Sync-Point Drop for Efficient Tensor Parallelism of Large Language Models
  6. ICML 2025The Panaceas for Improving Low-Rank Decomposition in Communication-Efficient Federated Learning
  7. ICML 2025Unified Breakdown Analysis for Byzantine Robust Gossip
  8. ICML 2025When, Where and Why to Average Weights?
  9. NeurIPS 2025A Fair Federated Learning Method for Handling Client Participation Probability Inconsistencies in Heterogeneous Environments
  10. NeurIPS 2025Accelerating Parallel Diffusion Model Serving with Residual Compression
  11. NeurIPS 2025ACCO: Accumulate While You Communicate for Communication-Overlapped Sharded LLM Training
  12. NeurIPS 2025Agentic Plan Caching: Test-Time Memory for Fast and Cost-Efficient LLM Agents
  13. NeurIPS 2025Approximate Gradient Coding for Distributed Learning with Heterogeneous Stragglers
  14. NeurIPS 2025AREAL: A Large-Scale Asynchronous Reinforcement Learning System for Language Reasoning
  15. NeurIPS 2025ASDSV: Multimodal Generation Made Efficient with Approximate Speculative Diffusion and Speculative Verification
  16. NeurIPS 2025Block-Diagonal LoRA for Eliminating Communication Overhead in Tensor Parallel LoRA Serving
  17. NeurIPS 2025Channel Simulation and Distributed Compression with Ensemble Rejection Sampling
  18. NeurIPS 2025ClusterFusion: Expanding Operator Fusion Scope for LLM Inference via Cluster-Level Collective Primitive
  19. NeurIPS 2025Communication-Efficient Diffusion Denoising Parallelization via Reuse-then-Predict Mechanism
  20. NeurIPS 2025Communication-Efficient Language Model Training Scales Reliably and Robustly: Scaling Laws for DiLoCo
  21. NeurIPS 2025Contextual Tokenization for Graph Inverted Indices
  22. NeurIPS 2025Cost-Efficient LLM Training with Lifetime-Aware Tensor Offloading via GPUDirect Storage
  23. NeurIPS 2025Distance Adaptive Beam Search for Provably Accurate Graph-Based Nearest Neighbor Search
  24. NeurIPS 2025Distributed mediation analysis with communication efficiency
  25. NeurIPS 2025Distributed Multi-Agent Bandits Over Erdős-Rényi Random Networks
  26. NeurIPS 2025DUO: No Compromise to Accuracy Degradation
  27. NeurIPS 2025DynaPipe: Dynamic Layer Redistribution for Efficient Serving of LLMs with Pipeline Parallelism
  28. NeurIPS 2025Efficient Pre-Training of LLMs via Topology-Aware Communication Alignment on More Than 9600 GPUs
  29. NeurIPS 2025Efficient Training-Free Online Routing for High-Volume Multi-LLM Serving
  30. NeurIPS 2025ElasticMM: Efficient Multimodal LLMs Serving with Elastic Multimodal Parallelism
  31. NeurIPS 2025EnCompass: Enhancing Agent Programming with Search Over Program Execution Paths
  32. NeurIPS 2025Exploiting Dynamic Sparsity in Einsum
  33. NeurIPS 2025FedLPA: Local Prior Alignment for Heterogeneous Federated Generalized Category Discovery
  34. NeurIPS 2025First Attentions Last: Better Exploiting First Attentions for Efficient Parallel Training
  35. NeurIPS 2025FlashMoE: Fast Distributed MoE in a Single Kernel
  36. NeurIPS 2025FlowMoE: A Scalable Pipeline Scheduling Framework for Distributed Mixture-of-Experts Training
  37. NeurIPS 2025Foresight: Adaptive Layer Reuse for Accelerated and High-Quality Text-to-Video Generation
  38. NeurIPS 2025Gatekeeper: Improving Model Cascades Through Confidence Tuning
  39. NeurIPS 2025Generative Caching for Structurally Similar Prompts and Responses
  40. NeurIPS 2025HyGen: Efficient LLM Serving via Elastic Online-Offline Request Co-location
  41. NeurIPS 2025InfantAgent-Next: A Multimodal Generalist Agent for Automated Computer Interaction
  42. NeurIPS 2025KVFlow: Efficient Prefix Caching for Accelerating LLM-Based Multi-Agent Workflows
  43. NeurIPS 2025Loquetier: A Virtualized Multi-LoRA Framework for Unified LLM Fine-tuning and Serving
  44. NeurIPS 2025MeCeFO: Enhancing LLM Training Robustness via Fault-Tolerant Optimization
  45. NeurIPS 2025Mixtures of Subspaces for Bandwidth Efficient Context Parallel Training
  46. NeurIPS 2025Mozart: Modularized and Efficient MoE Training on 3.5D Wafer-Scale Chiplet Architectures
  47. NeurIPS 2025msf-CNN: Patch-based Multi-Stage Fusion with Convolutional Neural Networks for TinyML
  48. NeurIPS 2025MUSTAFAR: Promoting Unstructured Sparsity for KV Cache Pruning in LLM Inference
  49. NeurIPS 2025NestedFP: High-Performance, Memory-Efficient Dual-Precision Floating Point Support for LLMs
  50. NeurIPS 2025PipeFusion: Patch-level Pipeline Parallelism for Diffusion Transformers Inference

在 Claude、Cursor 里直接问这个趋势

接上 Paperlist MCP 后，写综述、挑方向或准备投稿时，不用离开对话框就能拿到这页的数字：每年各会议的论文数、增长倍数、涨得最快的子方向和代表论文，AI 会按你的问题整理。

可以直接复制这句话问 AI

用 paperlist 查分布式训练在 NeurIPS 2023–2026 的论文趋势，哪一年涨得最快？

AI 会调用的工具：get_topic_stats {"topic": "distributed-training"}

两分钟接入 Paperlist MCP（Claude Code、Cursor、Codex、ChatGPT）

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。

「效率与系统」下的其他主题