高效训练论文趋势 2023–2026

Scaling law、MoE、蒸馏与训练效率

统计范围：NeurIPS、ICLR、ICML、CVPR 四个会议，2023–2026 年，共 1,765 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

高效训练论文数，按年份和会议
年份NeurIPSICLRICMLCVPR合计
2023112678384346
2024207812280300
202519012611882516
2026177148169109603

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

全部高效训练论文

年份全部2023202420252026
会议全部NeurIPSICLRICMLCVPR

共 190 篇，第 4 / 4 页

  1. NeurIPS 2025Revisiting Residual Connections: Orthogonal Updates for Stable and Efficient Deep Networks
  2. NeurIPS 2025Ridge Boosting is Both Robust and Efficient
  3. NeurIPS 2025Sample-efficient Learning of Concepts with Theoretical Guarantees: from Data to Concepts without Interventions
  4. NeurIPS 2025Sampling-Efficient Test-Time Scaling: Self-Estimating the Best-of-N Sampling in Early Decoding
  5. NeurIPS 2025Scaling Embedding Layers in Language Models
  6. NeurIPS 2025Scaling Law with Learning Rate Annealing
  7. NeurIPS 2025Scaling Up Liquid-Resistance Liquid-Capacitance Networks for Efficient Sequence Modeling
  8. NeurIPS 2025SDPGO: Efficient Self-Distillation Training Meets Proximal Gradient Optimization
  9. NeurIPS 2025Searching Efficient Semantic Segmentation Architectures via Dynamic Path Selection
  10. NeurIPS 2025SeerAttention: Self-distilled Attention Gating for Efficient Long-context Prefilling
  11. NeurIPS 2025Self-Refining Language Model Anonymizers via Adversarial Distillation
  12. NeurIPS 2025Sign-In to the Lottery: Reparameterizing Sparse Training
  13. NeurIPS 2025Simple and Efficient Heterogeneous Temporal Graph Neural Network
  14. NeurIPS 2025SmallKV: Small Model Assisted Compensation of KV Cache Compression for Efficient LLM Inference
  15. NeurIPS 2025Smooth Regularization for Efficient Video Recognition
  16. NeurIPS 2025SparseDiT: Token Sparsification for Efficient Diffusion Transformer
  17. NeurIPS 2025Spiking Meets Attention: Efficient Remote Sensing Image Super-Resolution with Attention Spiking Neural Networks
  18. NeurIPS 2025Spotlight Attention: Towards Efficient LLM Generation via Non-linear Hashing-based KV Cache Retrieval
  19. NeurIPS 2025Stackelberg Self-Annotation: A Robust Approach to Data-Efficient LLM Alignment
  20. NeurIPS 2025StreamBP: Memory-Efficient Exact Backpropagation for Long Sequence Training of LLMs
  21. NeurIPS 2025StreamForest: Efficient Online Video Understanding with Persistent Event Memory
  22. NeurIPS 2025Structured Sparse Transition Matrices to Enable State Tracking in State-Space Models
  23. NeurIPS 2025Subspace Networks: Scaling Decentralized Training with Communication-Efficient Model Parallelism
  24. NeurIPS 2025SubTrack++ : Gradient Subspace Tracking for Scalable LLM Training
  25. NeurIPS 2025SUMO: Subspace-Aware Moment-Orthogonalization for Accelerating Memory-Efficient LLM Training
  26. NeurIPS 2025Superposition Yields Robust Neural Scaling
  27. NeurIPS 2025SymMaP: Improving Computational Efficiency in Linear Solvers through Symbolic Preconditioning
  28. NeurIPS 2025Synergy Between the Strong and the Weak: Spiking Neural Networks are Inherently Self-Distillers
  29. NeurIPS 2025Teaching Transformers to Solve Combinatorial Problems through Efficient Trial & Error
  30. NeurIPS 2025Technical Debt in In-Context Learning: Diminishing Efficiency in Long Context
  31. NeurIPS 2025The Omni-Expert: A Computationally Efficient Approach to Achieve a Mixture of Experts in a Single Expert Model
  32. NeurIPS 2025Tiled Flash Linear Attention: More Efficient Linear RNN and xLSTM Kernels
  33. NeurIPS 2025Towards Interpretable and Efficient Attention: Compressing All by Contracting a Few
  34. NeurIPS 2025TrajMamba: An Efficient and Semantic-rich Vehicle Trajectory Pre-training Model
  35. NeurIPS 2025UMoE: Unifying Attention and FFN with Shared Experts
  36. NeurIPS 2025Unified Scaling Laws for Compressed Representations
  37. NeurIPS 2025V2V: Scaling Event-Based Vision through Efficient Video-to-Voxel Simulation
  38. NeurIPS 2025Venus-MAXWELL: Efficient Learning of Protein-Mutation Stability Landscapes using Protein Language Models
  39. NeurIPS 2025Zebra-Llama: Towards Extremely Efficient Hybrid Models
  40. NeurIPS 2025ZeroS: Zero‑Sum Linear Attention for Efficient Transformers

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 查高效训练在 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": "efficient-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高效训练论文？
共 1,765 篇，按年份分别是 2023 年 346 篇、2024 年 300 篇、2025 年 516 篇、2026 年 603 篇。
NeurIPS 2026 有多少篇高效训练论文？
NeurIPS 2026 有 177 篇高效训练论文，上一届是 190 篇。
ICLR 2026 有多少篇高效训练论文？
ICLR 2026 有 148 篇高效训练论文，上一届是 126 篇。
ICML 2026 有多少篇高效训练论文？
ICML 2026 有 169 篇高效训练论文，上一届是 118 篇。
CVPR 2026 有多少篇高效训练论文？
CVPR 2026 有 109 篇高效训练论文，上一届是 82 篇。
哪个会议的高效训练论文最多？
NeurIPS，2023–2026 年共 499 篇。

「效率与系统」下的其他主题