算子融合论文趋势 2023–2026

主要结论

• 算子融合论文从 2023 年的 51 篇降到 2026 年的 48 篇，热度在回落。
• NeurIPS 是发算子融合论文最多的会议，2023–2026 年共 73 篇，占四个会议总量的 33%。
• 交叉方向里涨得最快的是传感融合：同时属于两个主题的论文从 2023 年的 2 篇增至 2026 年的 12 篇。
• 在「硬件与编译」的 5 个主题里，它的 2026 年论文数排第 3；同组增长最快的端侧推理是 4.0 倍，算子融合是 0.9 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 2026Beyond Adjacent Layers: Graph-Guided Layer Fusion for Compressing Large Language Models
• ICML 2026Asymmetric conformal prediction with penalized kernel sum-of-squares
• ICLR 2026Probabilistic Kernel Function for Fast Angle Testing
• CVPR 2026Bridging RGB and Hematoxylin Components An Interleaved Guidance and Fusion
• NeurIPS 2025Beyond Benign Overfitting in Nadaraya-Watson Interpolators
• ICML 2025Equivariant Neural Tangent Kernels
• ICLR 2025DLEFT-MKC: Dynamic Late Fusion Multiple Kernel Clustering with Robust Tensor Learning via Min-Max Optimization
• CVPR 2025Parameterized Blur Kernel Prior Learning for Local Motion Deblurring
• NeurIPS 2024Efficient Convex Algorithms for Universal Kernel Learning
• ICML 2024Characterizing Overfitting in Kernel Ridgeless Regression Through the Eigenspectrum
• ICLR 2024Convolutional Deep Kernel Machines
• CVPR 2024Bi-SSC: Geometric-Semantic Bidirectional Fusion for Camera-based 3D Semantic Scene Completion
• NeurIPS 2023CosNet: A Generalized Spectral Kernel Network
• ICML 2023A Kernel Stein Test of Goodness of Fit for Sequential Models
• ICLR 2023A Kernel Perspective of Skip Connections in Convolutional Networks
• CVPR 2023Compacting Binary Neural Networks by Sparse Kernel Selection
• NeurIPS 2026Dynamical Adapter Fusion: Constructing A Global Adapter for Pre-Trained Model-based Class-Incremental Learning
• ICML 2026Even Faster Kernel Matrix Linear Algebra via Density Estimation
• ICLR 2026LaplacianFormer:Rethinking Linear Attention with Laplacian Kernel
• CVPR 2026CryoKRAQEN Kernel-Regularized Annealing for Quantized Embedding Networks in Cryo-EM Heterogeneous

全部算子融合论文

共 66 篇，第 1 / 2 页

1. ICML 2026Are Two Datasets Close Enough With Statistical Significance? A Kernel Distributional Closeness Testing Approach
2. ICML 2026Asymmetric conformal prediction with penalized kernel sum-of-squares
3. ICML 2026Dimensionality Reduction with Point-distributions Similarity Invariant
4. ICML 2026Diversity-Aware Recursive Feature Multiple Kernel Learning
5. ICML 2026Even Faster Kernel Matrix Linear Algebra via Density Estimation
6. ICML 2026Feasible Fusion: Constrained Joint Estimation under Structural Non-Overlap
7. ICML 2026Finite-Width Neural Tangent Kernels from Feynman Diagrams
8. ICML 2026Kernel-based Maximum-of-difference Test for Two-sample Comparison
9. ICML 2026Model Fusion via Retrofitting
10. ICML 2026New Bounds for Kernel Sums via Fast Spherical Embeddings
11. ICML 2026Preserving Plasticity in Continual Learning via Dynamical Isometry
12. ICML 2026Sequential Kernel-based Conditional Independence Testing via Adaptive Betting
13. ICML 2026UrbanFusion: Stochastic Multimodal Fusion for Contrastive Learning of Robust Spatial Representations
14. ICML 2025Adaptive kernel predictors from feature-learning infinite limits of neural networks
15. ICML 2025Beyond the Permutation Symmetry of Transformers: The Role of Rotation for Model Fusion
16. ICML 2025COKE: Core Kernel for More Efficient Approximation of Kernel Weights in Multiple Kernel Clustering
17. ICML 2025Equivariant Neural Tangent Kernels
18. ICML 2025K$^2$IE: Kernel Method-based Kernel Intensity Estimators for Inhomogeneous Poisson Processes
19. ICML 2025Learning Curves of Stochastic Gradient Descent in Kernel Regression
20. ICML 2025Learning Input Encodings for Kernel-Optimal Implicit Neural Representations
21. ICML 2025Nested Expectations with Kernel Quadrature
22. ICML 2025NTK-DFL: Enhancing Decentralized Federated Learning in Heterogeneous Settings via Neural Tangent Kernel
23. ICML 2025PEAKS: Selecting Key Training Examples Incrementally via Prediction Error Anchored by Kernel Similarity
24. ICML 2025Permutation-Free High-Order Interaction Tests
25. ICML 2025RollingQ: Reviving the Cooperation Dynamics in Multimodal Transformer
26. ICML 2025Scalable Gaussian Processes with Latent Kronecker Structure
27. ICML 2025The Price of Linear Time: Error Analysis of Structured Kernel Interpolation
28. ICML 2024Amortized Variational Deep Kernel Learning
29. ICML 2024Bring Your Own (Non-Robust) Algorithm to Solve Robust MDPs by Estimating The Worst Kernel
30. ICML 2024Characterizing Overfitting in Kernel Ridgeless Regression Through the Eigenspectrum
31. ICML 2024Copula-Nested Spectral Kernel Network
32. ICML 2024Deep Fusion: Efficient Network Training via Pre-trained Initializations
33. ICML 2024DeepPolar: Inventing Nonlinear Large-Kernel Polar Codes via Deep Learning
34. ICML 2024Dissecting Multimodality in VideoQA Transformer Models by Impairing Modality Fusion
35. ICML 2024Generalization in Kernel Regression Under Realistic Assumptions
36. ICML 2024High-Dimensional Kernel Methods under Covariate Shift: Data-Dependent Implicit Regularization
37. ICML 2024Is Kernel Prediction More Powerful than Gating in Convolutional Neural Networks?
38. ICML 2024Kernel Debiased Plug-in Estimation: Simultaneous, Automated Debiasing without Influence Functions for Many Target Parameters
39. ICML 2024Kernel Semi-Implicit Variational Inference
40. ICML 2024Kernel-Based Evaluation of Conditional Biological Sequence Models
41. ICML 2024Mean-field Analysis on Two-layer Neural Networks from a Kernel Perspective
42. ICML 2024Neural-Kernel Conditional Mean Embeddings
43. ICML 2024On the Consistency of Kernel Methods with Dependent Observations
44. ICML 2024Optimal Kernel Quantile Learning with Random Features
45. ICML 2024Position: Opportunities Exist for Machine Learning in Magnetic Fusion Energy
46. ICML 2024Predictive Dynamic Fusion
47. ICML 2024Quasi-Monte Carlo Features for Kernel Approximation
48. ICML 2024Revealing the Dark Secrets of Extremely Large Kernel ConvNets on Robustness
49. ICML 2024Reward-Free Kernel-Based Reinforcement Learning
50. ICML 2024Sampling in Unit Time with Kernel Fisher-Rao Flow

常见问题

2023–2026 年四大会议一共有多少篇算子融合论文？

共 221 篇，按年份分别是 2023 年 51 篇、2024 年 66 篇、2025 年 56 篇、2026 年 48 篇。

NeurIPS 2026 有多少篇算子融合论文？

NeurIPS 2026 有 14 篇算子融合论文，上一届是 25 篇。

ICLR 2026 有多少篇算子融合论文？

ICLR 2026 有 9 篇算子融合论文，上一届是 8 篇。

ICML 2026 有多少篇算子融合论文？

ICML 2026 有 13 篇算子融合论文，上一届是 14 篇。

CVPR 2026 有多少篇算子融合论文？

CVPR 2026 有 12 篇算子融合论文，上一届是 9 篇。

哪个会议的算子融合论文最多？

NeurIPS，2023–2026 年共 73 篇。