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 1,765 篇，第 3 / 36 页

101. CVPR 2026Trainable Log-linear Sparse Attention for Efficient Diffusion Transformers
102. CVPR 2026TriLite Efficient Weakly Supervised Object Localization with Universal Visual Features
103. CVPR 2026UETrack A Unified and Efficient Framework for Single Object Tracking
104. CVPR 2026VDOT Efficient Unified Video Creation via Optimal Transport Distillation
105. CVPR 2026VEMamba Efficient Isotropic Reconstruction of Volume Electron Microscopy with Axial-Lateral
106. CVPR 2026VMonarch Efficient Video Diffusion Transformers with Structured Attention
107. CVPR 2026WhisperNet A Scalable Solution for Bandwidth-Efficient Collaboration
108. CVPR 2026WikiCLIP An Efficient Contrastive Baseline for Open-domain Visual Entity Recognition
109. CVPR 2026YOSE You Only Select Essential Tokens for Efficient DiT-based Video
110. ICLR 2026$\mu$LO: Compute-Efficient Meta-Generalization of Learned Optimizers
111. ICLR 2026A Generalized Geometric Theoretical Framework of Centroid Discriminant Analysis for Linear Classification of Multi-dimensional Data
112. ICLR 2026A universal compression theory: Lottery ticket hypothesis and superpolynomial scaling laws
113. ICLR 2026Achieving low-bit Muon through subspace preservation and grid quantization
114. ICLR 2026An Efficient SE(p)-Invariant Transport Metric Driven by Polar Transport Discrepancy-based Representation
115. ICLR 2026Annotation-Efficient Honesty Alignment via Confidence Elicitation and Calibration
116. ICLR 2026AnyBCQ: Hardware Efficient Flexible Binary-Coded Quantization for Multi-Precision LLMs
117. ICLR 2026AQER: A Scalable and Efficient Data Loader for Digital Quantum Computers
118. ICLR 2026ATLAS: Adaptive Transfer Scaling Laws for Multilingual Pretraining, Finetuning, and Decoding the Curse of Multilinguality
119. ICLR 2026Batch Pruning by Activation Stability
120. ICLR 2026Beyond Masks: Efficient, Flexible Diffusion Language Models via Deletion-Insertion Processes
121. ICLR 2026Beyond Outliers: A Study of Optimizers Under Quantization
122. ICLR 2026Beyond Student: An Asymmetric Network for Neural Network Inheritance
123. ICLR 2026Beyond URLs: Metadata Diversity and Position for Efficient LLM Pretraining
124. ICLR 2026BLADE: Block-Sparse Attention Meets Step Distillation for Efficient Video Generation
125. ICLR 2026Break the Trade-off Between Watermark Strength and Speculative Sampling Efficiency for Language Models
126. ICLR 2026Cascadia: An Efficient Cascade Serving System for Large Language Models
127. ICLR 2026CaTS: Calibrated Test-Time Scaling for Efficient LLM Reasoning
128. ICLR 2026Channel-Aware Mixed-Precision Quantization for Efficient Long-Context Inference
129. ICLR 2026Composer: A Search Framework for Hybrid Neural Architecture Design
130. ICLR 2026Compute-Optimal Quantization-Aware Training
131. ICLR 2026Convex Efficient Coding
132. ICLR 2026Cost-Aware Dynamic Tree Construction for Efficient Large Language Model Inference
133. ICLR 2026CR-Net: Scaling Parameter-Efficient Training with Cross-Layer Low-Rank Structure
134. ICLR 2026DeCo-DETR: Decoupled Cognition DETR for efficient Open-Vocabulary Object Detection
135. ICLR 2026Difficulty–Diversity Collaborative Filtering for Data-Efficient LLM Fine-Tuning
136. ICLR 2026DiffusionBlocks: Block-wise Neural Network Training via Diffusion Interpretation
137. ICLR 2026DisTaC: Conditioning Task Vectors via Distillation for Robust Model Merging
138. ICLR 2026Distillation of Large Language Models via Concrete Score Matching
139. ICLR 2026Distilling to Hybrid Attention Models via KL-Guided Layer Selection
140. ICLR 2026DPad: Efficient Diffusion Language Models with Suffix Dropout
141. ICLR 2026DPQuant: Efficient and Private Model Training via Dynamic Quantization Scheduling
142. ICLR 2026Dual Language Models: Balancing sample-efficiency and overfitting resilience
143. ICLR 2026DUET: DISTILLED LLM UNLEARNING FROM AN EFFICIENTLY CONTEXTUALIZED TEACHER
144. ICLR 2026Efficient and Sharp Off-Policy Learning under Unobserved Confounding
145. ICLR 2026Efficient Approximate Posterior Sampling with Annealed Langevin Monte Carlo
146. ICLR 2026Efficient Autoregressive Inference for Transformer Probabilistic Models
147. ICLR 2026Efficient Best-of-Both-Worlds Algorithms for Contextual Combinatorial Semi-Bandits
148. ICLR 2026Efficient Credal Prediction through Decalibration
149. ICLR 2026Efficient Degradation-agnostic Image Restoration via Channel-Wise Functional Decomposition and Manifold Regularization
150. ICLR 2026Efficient Differentiable Contact Model with Long-range Influence

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。