高效训练论文趋势 2023–2026

主要结论

• 高效训练论文从 2023 年的 346 篇增长到 2026 年的 603 篇，三年间扩大到约 1.7 倍。
• NeurIPS 是发高效训练论文最多的会议，2023–2026 年共 499 篇，占四个会议总量的 28%。
• 交叉方向里涨得最快的是深度学习 · large language models：同时属于两个主题的论文从 2023 年的 6 篇增至 2026 年的 50 篇。
• 在「效率与系统」的 5 个主题里，它的 2026 年论文数排第 3；同组增长最快的硬件协同是 4.4 倍，高效训练是 1.7 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 2026Adaptive Residual Quantization for Memory-Efficient Temporal Action Segmentation
• ICML 2026A Geometry-Aware Efficient Algorithm for Compositional Entropic Risk Minimization
• ICLR 2026An Efficient SE(p)-Invariant Transport Metric Driven by Polar Transport Discrepancy-based Representation
• CVPR 2026A Bit is All You Need Efficient Video Capture via
• NeurIPS 20254DGCPro: Efficient Hierarchical 4D Gaussian Compression for Progressive Volumetric Video Streaming
• ICML 2025A Sample Efficient Conditional Independence Test in the Presence of Discretization
• ICLR 2025Breaking Class Barriers: Efficient Dataset Distillation via Inter-Class Feature Compensator
• CVPR 2025Accurate Scene Text Recognition with Efficient Model Scaling and Cloze Self-Distillation
• NeurIPS 2024$\texttt{dattri}$: A Library for Efficient Data Attribution
• ICML 2024A fast algorithm to simulate nonlinear resistive networks
• ICLR 2024Dictionary Contrastive Learning for Efficient Local Supervision without Auxiliary Networks
• CVPR 2024Deep-TROJ: An Inference Stage Trojan Insertion Algorithm through Efficient Weight Replacement Attack
• NeurIPS 2023Alternating Updates for Efficient Transformers
• ICML 2023Auxiliary Modality Learning with Generalized Curriculum Distillation
• ICLR 2023A Model or 603 Exemplars: Towards Memory-Efficient Class-Incremental Learning
• CVPR 2023A Rotation-Translation-Decoupled Solution for Robust and Efficient Visual-Inertial Initialization
• NeurIPS 2026Advanced Routing as Regularization Allocation for Efficient Diffusion Transformer Training
• ICML 2026Active Learning with Foundation Model Priors: Efficient Learning under Class Imbalance
• ICLR 2026AQER: A Scalable and Efficient Data Loader for Digital Quantum Computers
• CVPR 2026Adaptive Bayesian Early-Exit Networks for Efficient Non-Transferable Learning

全部高效训练论文

共 603 篇，第 9 / 13 页

401. ICML 2026Sparser, Faster, Lighter Transformer Language Models
402. ICML 2026SparseSSM: Efficient Selective Structured State Space Models Can Be Pruned in One-Shot
403. ICML 2026Spatial-Aware Reduction Framework: Towards Efficient and Faithful Visual State Space Models
404. ICML 2026SpecForge: A Flexible and Efficient Open-Source Training Framework for Speculative Decoding
405. ICML 2026Spectral Reach: Understanding Neural Scaling as Progress into the Spectral Tail
406. ICML 2026Spik4lite: Refactoring Neuromorphic Sparsity for Efficient Spiking Neural Networks on Commodity Edge Devices
407. ICML 2026SSA: Sparse Sparse Attention by Aligning Full and Sparse Attention Outputs in Feature Space
408. ICML 2026Swordsman: Entropy-Driven Adaptive Block Partition for Efficient Diffusion Language Models
409. ICML 2026TabPack: Efficient Hyperparameter Ensembles for Tabular Deep Learning
410. ICML 2026TetraJet-v2: Accurate NVFP4 Training for Large Language Models with Oscillation Suppression and Outlier Control
411. ICML 2026ToaSt: Token Channel Selection and Structured Pruning for Efficient ViT
412. ICML 2026Token Sparse Attention: Efficient Long-Context Inference with Interleaved Token Selection
413. ICML 2026TokenDrop: Token-Level Importance-Aware Backward Propagation Skipping for Efficient LLM Fine-Tuning
414. ICML 2026Towards Resource-Efficient LLMs: End-to-End Energy Accounting of Distillation Pipelines
415. ICML 2026Training-Free Hashing-Based Attention via Binary Principal Components
416. ICML 2026Turning Back Without Forgetting: Selective Backward Refinement for Parameter-Efficient Continual Learning
417. ICML 2026Turning Stale Gradients into Stable Gradients: Coherent Coordinate Descent with Implicit Landscape Smoothing for Lightweight Zeroth-Order Optimization
418. ICML 2026UniFast-HGR: Scalable and Efficient Maximal Correlation for Multimodal Models
419. ICML 2026Unifying Dataset Pruning and Distillation for Efficient Large-scale Compression
420. ICML 2026Unifying Stacking and Cascading for Efficient Ensemble Inference
421. ICML 2026Universal One-third Time Scaling in Learning Peaked Distributions
422. ICML 2026Untied Ulysses: Memory-Efficient Context Parallelism via Headwise Chunking
423. ICML 2026Utility Boundary of Dataset Distillation: Scaling and Coverage Laws
424. ICML 2026WAVE: Window-Aware Vocabulary-Efficient Early-Exit for Training-Free LLM Acceleration
425. ICML 2026WBMM: Windowed Batch Matrix Multiplication for Efficient Large Receptive Field Convolution
426. ICML 2026ZipMoE: Efficient On-Device MoE Serving via Lossless Compression and Cache-Affinity Scheduling
427. NeurIPS 2026$\delta$-Mem: : Efficient Online Memory For Large Language Models
428. NeurIPS 2026A Grid Efficient Transport Optimization Model for the Wasserstein Barycenter Problem
429. NeurIPS 2026A Solver-Efficient Neural Adversarial Attack on Subgraph Matching Models
430. NeurIPS 2026Achieve Latency-Efficient Temporal-Coding Spiking LLMs via Discretization-Aware Conversion
431. NeurIPS 2026Adapting in the Dark: Efficient and Stable Test-Time Adaptation for Black-Box Models
432. NeurIPS 2026Adaptive Residual Quantization for Memory-Efficient Temporal Action Segmentation
433. NeurIPS 2026ADIS-Law: Unified Scaling Laws for Annealing-Phase Domain Injection in Large Language Models
434. NeurIPS 2026Advanced Routing as Regularization Allocation for Efficient Diffusion Transformer Training
435. NeurIPS 2026AgentWeave: Efficient Distributed Agent Serving via Flow Decomposition
436. NeurIPS 2026An Efficient Cross-modal Feature Reconstruction Model for Multimodal Multi-class Anomaly Detection
437. NeurIPS 2026ASVQ: What Reparameterization Is Just Enough for Efficient Codebook Learning?
438. NeurIPS 2026Asymmetric Scaling Laws from Sparse Features
439. NeurIPS 2026Attribute-Efficient Learning of Sparse Halfspaces with Constant Malicious Noise Rate
440. NeurIPS 2026Benchmarks as Measurement Instruments: Quantifying Signal and Noise for More Efficient AI Evaluations Under Distribution Shift
441. NeurIPS 2026Beyond Pairwise Supervision: Spectral Characteristic Matching for Data-Efficient Multimodal Alignment
442. NeurIPS 2026Beyond the Half Approximation: Fair and Efficient Online Class Matching
443. NeurIPS 2026Bidirectional Information Flow (BIF) - A Sample Efficient Hierarchical Gaussian Process for Bayesian Optimization
444. NeurIPS 2026BiHashFormer: Hash-Driven Dual-Branch Transformer for Efficient Object Detection in HRW Shots
445. NeurIPS 2026BLANP: Memory-Efficient Backpropagation-Free Local Training via Antithetic Node Perturbation
446. NeurIPS 2026BROS: Bias-Corrected Randomized Subspaces for Memory-Efficient Single-Loop Bilevel Optimization
447. NeurIPS 2026ByteDistill: Cross-Tokenizer Distillation via Chunk-wise Byte-Level Distribution Alignment
448. NeurIPS 2026CELEUS: Certifiable and Efficient LLM Evaluation via E-Processes
449. NeurIPS 2026CHoRD: Coordinating Scheduling and Data Placement for Efficient Deep Neural Network Inference on Chiplet-Based GPUs
450. NeurIPS 2026Communication-Efficient LLM Adaptation over Decentralized GPU Meshes

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。