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 516 篇，第 10 / 11 页

  1. NeurIPS 2025Mixtures of Subspaces for Bandwidth Efficient Context Parallel Training
  2. NeurIPS 2025MoESD: Unveil Speculative Decoding's Potential for Accelerating Sparse MoE
  3. NeurIPS 2025Mozart: Modularized and Efficient MoE Training on 3.5D Wafer-Scale Chiplet Architectures
  4. NeurIPS 2025Nemotron-Flash: Towards Latency-Optimal Hybrid Small Language Models
  5. NeurIPS 2025Noise-Robustness Through Noise: A Framework combining Asymmetric LoRA with Poisoning MoE
  6. NeurIPS 2025Oracle-Efficient Combinatorial Semi-Bandits
  7. NeurIPS 2025Pan-LUT: Efficient Pan-sharpening via Learnable Look-Up Tables
  8. NeurIPS 2025Parallel Scaling Law for Language Models
  9. NeurIPS 2025ParetoQ: Improving Scaling Laws in Extremely Low-bit LLM Quantization
  10. NeurIPS 2025Pay Attention to Small Weights
  11. NeurIPS 2025Position: AI Should Sense Better, Not Just Scale Bigger: Adaptive Sensing as a Paradigm Shift
  12. NeurIPS 2025Power Lines: Scaling laws for weight decay and batch size in LLM pre-training
  13. NeurIPS 2025PREAMBLE: Private and Efficient Aggregation via Block Sparse Vectors
  14. NeurIPS 2025Predictable Scale (Part II) --- Farseer: A Refined Scaling Law in LLMs
  15. NeurIPS 2025Private Statistical Estimation via Truncation
  16. NeurIPS 2025Private Training Large-scale Models with Efficient DP-SGD
  17. NeurIPS 2025Pruning-Robust Mamba with Asymmetric Multi-Scale Scanning Paths
  18. NeurIPS 2025PT-MoE: An Efficient Finetuning Framework for Integrating Mixture-of-Experts into Prompt Tuning
  19. NeurIPS 2025Q-Palette: Fractional-Bit Quantizers Toward Optimal Bit Allocation for Efficient LLM Deployment
  20. NeurIPS 2025Quadratic Coreset Selection: Certifying and Reconciling Sequence and Token Mining for Efficient Instruction Tuning
  21. NeurIPS 2025Random Search Neural Networks for Efficient and Expressive Graph Learning
  22. NeurIPS 2025RankSEG-RMA: An Efficient Segmentation Algorithm via Reciprocal Moment Approximation
  23. NeurIPS 2025RCCDA: Adaptive Model Updates in the Presence of Concept Drift under a Constrained Resource Budget
  24. NeurIPS 2025REN: Fast and Efficient Region Encodings from Patch-Based Image Encoders
  25. NeurIPS 2025REP: Resource-Efficient Prompting for Rehearsal-Free Continual Learning
  26. NeurIPS 2025Rethinking Optimal Verification Granularity for Compute-Efficient Test-Time Scaling
  27. NeurIPS 2025Revisiting Residual Connections: Orthogonal Updates for Stable and Efficient Deep Networks
  28. NeurIPS 2025Ridge Boosting is Both Robust and Efficient
  29. NeurIPS 2025Sample-efficient Learning of Concepts with Theoretical Guarantees: from Data to Concepts without Interventions
  30. NeurIPS 2025Sampling-Efficient Test-Time Scaling: Self-Estimating the Best-of-N Sampling in Early Decoding
  31. NeurIPS 2025Scaling Embedding Layers in Language Models
  32. NeurIPS 2025Scaling Law with Learning Rate Annealing
  33. NeurIPS 2025Scaling Up Liquid-Resistance Liquid-Capacitance Networks for Efficient Sequence Modeling
  34. NeurIPS 2025SDPGO: Efficient Self-Distillation Training Meets Proximal Gradient Optimization
  35. NeurIPS 2025Searching Efficient Semantic Segmentation Architectures via Dynamic Path Selection
  36. NeurIPS 2025SeerAttention: Self-distilled Attention Gating for Efficient Long-context Prefilling
  37. NeurIPS 2025Self-Refining Language Model Anonymizers via Adversarial Distillation
  38. NeurIPS 2025Sign-In to the Lottery: Reparameterizing Sparse Training
  39. NeurIPS 2025Simple and Efficient Heterogeneous Temporal Graph Neural Network
  40. NeurIPS 2025SmallKV: Small Model Assisted Compensation of KV Cache Compression for Efficient LLM Inference
  41. NeurIPS 2025Smooth Regularization for Efficient Video Recognition
  42. NeurIPS 2025SparseDiT: Token Sparsification for Efficient Diffusion Transformer
  43. NeurIPS 2025Spiking Meets Attention: Efficient Remote Sensing Image Super-Resolution with Attention Spiking Neural Networks
  44. NeurIPS 2025Spotlight Attention: Towards Efficient LLM Generation via Non-linear Hashing-based KV Cache Retrieval
  45. NeurIPS 2025Stackelberg Self-Annotation: A Robust Approach to Data-Efficient LLM Alignment
  46. NeurIPS 2025StreamBP: Memory-Efficient Exact Backpropagation for Long Sequence Training of LLMs
  47. NeurIPS 2025StreamForest: Efficient Online Video Understanding with Persistent Event Memory
  48. NeurIPS 2025Structured Sparse Transition Matrices to Enable State Tracking in State-Space Models
  49. NeurIPS 2025Subspace Networks: Scaling Decentralized Training with Communication-Efficient Model Parallelism
  50. NeurIPS 2025SubTrack++ : Gradient Subspace Tracking for Scalable LLM Training

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