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 346 篇，第 6 / 7 页

  1. NeurIPS 2023Birder: Communication-Efficient 1-bit Adaptive Optimizer for Practical Distributed DNN Training
  2. NeurIPS 2023Cal-QL: Calibrated Offline RL Pre-Training for Efficient Online Fine-Tuning
  3. NeurIPS 2023Clustering the Sketch: Dynamic Compression for Embedding Tables
  4. NeurIPS 2023Composing Parameter-Efficient Modules with Arithmetic Operation
  5. NeurIPS 2023Connecting Certified and Adversarial Training
  6. NeurIPS 2023Coordinating Distributed Example Orders for Provably Accelerated Training
  7. NeurIPS 2023CoPriv: Network/Protocol Co-Optimization for Communication-Efficient Private Inference
  8. NeurIPS 2023Data Portraits: Recording Foundation Model Training Data
  9. NeurIPS 2023Data Pruning via Moving-one-Sample-out
  10. NeurIPS 2023Defending against Data-Free Model Extraction by Distributionally Robust Defensive Training
  11. NeurIPS 2023DESSERT: An Efficient Algorithm for Vector Set Search with Vector Set Queries
  12. NeurIPS 2023Direct Training of SNN using Local Zeroth Order Method
  13. NeurIPS 2023Distributionally Robust Ensemble of Lottery Tickets Towards Calibrated Sparse Network Training
  14. NeurIPS 2023Does a sparse ReLU network training problem always admit an optimum ?
  15. NeurIPS 2023Don’t just prune by magnitude! Your mask topology is a secret weapon
  16. NeurIPS 2023DoWG Unleashed: An Efficient Universal Parameter-Free Gradient Descent Method
  17. NeurIPS 2023Dynamic Context Pruning for Efficient and Interpretable Autoregressive Transformers
  18. NeurIPS 2023Efficient Activation Function Optimization through Surrogate Modeling
  19. NeurIPS 2023Efficient Adversarial Contrastive Learning via Robustness-Aware Coreset Selection
  20. NeurIPS 2023Efficient Algorithms for Generalized Linear Bandits with Heavy-tailed Rewards
  21. NeurIPS 2023Efficient Batched Algorithm for Contextual Linear Bandits with Large Action Space via Soft Elimination
  22. NeurIPS 2023Efficient Bayesian Learning Curve Extrapolation using Prior-Data Fitted Networks
  23. NeurIPS 2023Efficient Data Subset Selection to Generalize Training Across Models: Transductive and Inductive Networks
  24. NeurIPS 2023Efficient Learning of Linear Graph Neural Networks via Node Subsampling
  25. NeurIPS 2023Efficient Online Clustering with Moving Costs
  26. NeurIPS 2023Efficient Robust Bayesian Optimization for Arbitrary Uncertain inputs
  27. NeurIPS 2023Efficient Subgame Refinement for Extensive-form Games
  28. NeurIPS 2023Efficient Testable Learning of Halfspaces with Adversarial Label Noise
  29. NeurIPS 2023Efficient Training of Energy-Based Models Using Jarzynski Equality
  30. NeurIPS 2023EgoDistill: Egocentric Head Motion Distillation for Efficient Video Understanding
  31. NeurIPS 2023Eliminating Catastrophic Overfitting Via Abnormal Adversarial Examples Regularization
  32. NeurIPS 2023Energy-Efficient Scheduling with Predictions
  33. NeurIPS 2023Explainable and Efficient Randomized Voting Rules
  34. NeurIPS 2023Exploring Loss Functions for Time-based Training Strategy in Spiking Neural Networks
  35. NeurIPS 2023FiGURe: Simple and Efficient Unsupervised Node Representations with Filter Augmentations
  36. NeurIPS 2023Generalizable Lightweight Proxy for Robust NAS against Diverse Perturbations
  37. NeurIPS 2023GradOrth: A Simple yet Efficient Out-of-Distribution Detection with Orthogonal Projection of Gradients
  38. NeurIPS 2023H3T: Efficient Integration of Memory Optimization and Parallelism for Large-scale Transformer Training
  39. NeurIPS 2023How2comm: Communication-Efficient and Collaboration-Pragmatic Multi-Agent Perception
  40. NeurIPS 2023Hybrid Search for Efficient Planning with Completeness Guarantees
  41. NeurIPS 2023Label Robust and Differentially Private Linear Regression: Computational and Statistical Efficiency
  42. NeurIPS 2023Language Semantic Graph Guided Data-Efficient Learning
  43. NeurIPS 2023Learning Efficient Coding of Natural Images with Maximum Manifold Capacity Representations
  44. NeurIPS 2023Mask Propagation for Efficient Video Semantic Segmentation
  45. NeurIPS 2023Maximum Independent Set: Self-Training through Dynamic Programming
  46. NeurIPS 2023MEGABYTE: Predicting Million-byte Sequences with Multiscale Transformers
  47. NeurIPS 2023Memory Efficient Optimizers with 4-bit States
  48. NeurIPS 2023MGDD: A Meta Generator for Fast Dataset Distillation
  49. NeurIPS 2023MixFormerV2: Efficient Fully Transformer Tracking
  50. NeurIPS 2023MosaicBERT: A Bidirectional Encoder Optimized for Fast Pretraining

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