混合专家论文趋势 2023–2026

主要结论

• 混合专家论文从 2023 年的 236 篇增长到 2026 年的 326 篇，三年间扩大到约 1.4 倍。
• NeurIPS 是发混合专家论文最多的会议，2023–2026 年共 354 篇，占四个会议总量的 39%。
• 交叉方向里涨得最快的是多模态大模型：同时属于两个主题的论文从 2023 年的 4 篇增至 2026 年的 79 篇。
• 在「优化与缩放」的 5 个主题里，它的 2026 年论文数排第 4；同组增长最快的缩放定律是 4.1 倍，混合专家是 1.4 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 2026A Biconvex Formulation for Stable Transport of Mixture Models with a Unique Solution
• ICML 2026$\phi$-Balancing for Mixture-of-Experts Training
• ICLR 2026A Federated Generalized Expectation-Maximization Algorithm for Mixture Models with an Unknown Number of Components
• CVPR 2026Divide Conquer and Aggregate Asymmetric Experts for Class-Imbalanced Semi-Supervised Medical
• NeurIPS 2025Advancing Expert Specialization for Better MoE
• ICML 2025$\texttt{I$^2$MoE}$: Interpretable Multimodal Interaction-aware Mixture-of-Experts
• ICLR 2025A Sanity Check for AI-generated Image Detection
• CVPR 2025Complexity Experts are Task-Discriminative Learners for Any Image Restoration
• NeurIPS 2024$\texttt{Model-GLUE}$: Democratized LLM Scaling for A Large Model Zoo in the Wild
• ICML 2024$\texttt{MoE-RBench}$: Towards Building Reliable Language Models with Sparse Mixture-of-Experts
• ICLR 2024Fusing Models with Complementary Expertise
• CVPR 2024MoDE: CLIP Data Experts via Clustering
• NeurIPS 2023[Re] Numerical influence of ReLU'(0) on backpropagation
• ICML 2023A Distribution Optimization Framework for Confidence Bounds of Risk Measures
• ICLR 2023Knowledge-in-Context: Towards Knowledgeable Semi-Parametric Language Models
• CVPR 2023Balanced Product of Calibrated Experts for Long-Tailed Recognition
• NeurIPS 2026AdERA: Adaptive Exponent Reuse for Lossless Allgather in Sharded MoE Training
• ICML 2026Attention Sink Forges Native MoE in Attention Layers: Sink-Aware Training to Address Head Collapse
• ICLR 2026Abstracting Robot Manipulation Skills via Mixture-of-Experts Diffusion Policies
• CVPR 2026ERMoE Eigen-Reparameterized Mixture-of-Experts for Stable Routing and Interpretable Specialization

全部混合专家论文

共 172 篇，第 3 / 4 页

101. ICLR 2025Many-Objective Multi-Solution Transport
102. ICLR 2025Mixture Compressor for Mixture-of-Experts LLMs Gains More
103. ICLR 2025Mixture of Attentions For Speculative Decoding
104. ICLR 2025Mixture of Experts Made Personalized: Federated Prompt Learning for Vision-Language Models
105. ICLR 2025Mixture of In-Context Prompters for Tabular PFNs
106. ICLR 2025Mixture of Parrots: Experts improve memorization more than reasoning
107. ICLR 2025Mixture-of-Agents Enhances Large Language Model Capabilities
108. ICLR 2025MoE++: Accelerating Mixture-of-Experts Methods with Zero-Computation Experts
109. ICLR 2025MoLEx: Mixture of Layer Experts for Fine-tuning with Sparse Upcycling
110. ICLR 2025Monet: Mixture of Monosemantic Experts for Transformers
111. ICLR 2025More Experts Than Galaxies: Conditionally-Overlapping Experts with Biologically-Inspired Fixed Routing
112. ICLR 2025MoS: Unleashing Parameter Efficiency of Low-Rank Adaptation with Mixture of Shards
113. ICLR 2025Multiple Heads are Better than One: Mixture of Modality Knowledge Experts for Entity Representation Learning
114. ICLR 2025NetMoE: Accelerating MoE Training through Dynamic Sample Placement
115. ICLR 2025No Need to Talk: Asynchronous Mixture of Language Models
116. ICLR 2025OLMoE: Open Mixture-of-Experts Language Models
117. ICLR 2025pMoE: Prompting Diverse Experts Together Wins More in Visual Adaptation
118. ICLR 2025Provable Uncertainty Decomposition via Higher-Order Calibration
119. ICLR 2025RegMix: Data Mixture as Regression for Language Model Pre-training
120. ICLR 2025ReMoE: Fully Differentiable Mixture-of-Experts with ReLU Routing
121. ICLR 2025Revisiting Prefix-tuning: Statistical Benefits of Reparameterization among Prompts
122. ICLR 2025Routing Experts: Learning to Route Dynamic Experts in Existing Multi-modal Large Language Models
123. ICLR 2025Self-MoE: Towards Compositional Large Language Models with Self-Specialized Experts
124. ICLR 2025Solving Token Gradient Conflict in Mixture-of-Experts for Large Vision-Language Model
125. ICLR 2025Statistical Advantages of Perturbing Cosine Router in Mixture of Experts
126. ICLR 2025TC-MoE: Augmenting Mixture of Experts with Ternary Expert Choice
127. ICLR 2025Theory on Mixture-of-Experts in Continual Learning
128. ICLR 2025Tight Clusters Make Specialized Experts
129. ICLR 2025Time-MoE: Billion-Scale Time Series Foundation Models with Mixture of Experts
130. ICLR 2025Transformers Provably Learn Two-Mixture of Linear Classification via Gradient Flow
131. ICLR 2025Warm Diffusion: Recipe for Blur-Noise Mixture Diffusion Models
132. ICLR 2025Your Mixture-of-Experts LLM Is Secretly an Embedding Model for Free
133. ICLR 2024Chain-of-Experts: When LLMs Meet Complex Operations Research Problems
134. ICLR 2024Divide and not forget: Ensemble of selectively trained experts in Continual Learning
135. ICLR 2024From Sparse to Soft Mixtures of Experts
136. ICLR 2024Fusing Models with Complementary Expertise
137. ICLR 2024High-dimensional SGD aligns with emerging outlier eigenspaces
138. ICLR 2024Horizon-free Reinforcement Learning in Adversarial Linear Mixture MDPs
139. ICLR 2024Hybrid Sharing for Multi-Label Image Classification
140. ICLR 2024Large Language Model Cascades with Mixture of Thought Representations for Cost-Efficient Reasoning
141. ICLR 2024Learning with Mixture of Prototypes for Out-of-Distribution Detection
142. ICLR 2024M3C: A Framework towards Convergent, Flexible, and Unsupervised Learning of Mixture Graph Matching and Clustering
143. ICLR 2024Merge, Then Compress: Demystify Efficient SMoE with Hints from Its Routing Policy
144. ICLR 2024Mixture of LoRA Experts
145. ICLR 2024Mixture of Weak and Strong Experts on Graphs
146. ICLR 2024Mixture-of-Experts Meets Instruction Tuning: A Winning Combination for Large Language Models
147. ICLR 2024Object-Centric Learning with Slot Mixture Module
148. ICLR 2024Octavius: Mitigating Task Interference in MLLMs via LoRA-MoE
149. ICLR 2024Pushing Mixture of Experts to the Limit: Extremely Parameter Efficient MoE for Instruction Tuning
150. ICLR 2024Scaling physics-informed hard constraints with mixture-of-experts

常见问题

2023–2026 年四大会议一共有多少篇混合专家论文？

共 904 篇，按年份分别是 2023 年 236 篇、2024 年 119 篇、2025 年 223 篇、2026 年 326 篇。

NeurIPS 2026 有多少篇混合专家论文？

NeurIPS 2026 有 75 篇混合专家论文，上一届是 86 篇。

ICLR 2026 有多少篇混合专家论文？

ICLR 2026 有 73 篇混合专家论文，上一届是 59 篇。

ICML 2026 有多少篇混合专家论文？

ICML 2026 有 132 篇混合专家论文，上一届是 57 篇。

CVPR 2026 有多少篇混合专家论文？

CVPR 2026 有 46 篇混合专家论文，上一届是 21 篇。

哪个会议的混合专家论文最多？

NeurIPS，2023–2026 年共 354 篇。