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 56 篇，第 1 / 2 页

1. CVPR 2025Deformable Radial Kernel Splatting
2. CVPR 2025Layered Motion Fusion: Lifting Motion Segmentation to 3D in Egocentric Videos
3. CVPR 2025Parameterized Blur Kernel Prior Learning for Local Motion Deblurring
4. CVPR 2025RaCFormer: Towards High-Quality 3D Object Detection via Query-based Radar-Camera Fusion
5. CVPR 2025SGFormer: Satellite-Ground Fusion for 3D Semantic Scene Completion
6. CVPR 2025ShiftwiseConv: Small Convolutional Kernel with Large Kernel Effect
7. CVPR 2025SVDC: Consistent Direct Time-of-Flight Video Depth Completion with Frequency Selective Fusion
8. CVPR 2025TacoDepth: Towards Efficient Radar-Camera Depth Estimation with One-stage Fusion
9. CVPR 2025UltraFusion: Ultra High Dynamic Imaging using Exposure Fusion
10. ICLR 2025Coreset Spectral Clustering
11. ICLR 2025Deep Kernel Posterior Learning under Infinite Variance Prior Weights
12. ICLR 2025Divergence of Neural Tangent Kernel in Classification Problems
13. ICLR 2025DLEFT-MKC: Dynamic Late Fusion Multiple Kernel Clustering with Robust Tensor Learning via Min-Max Optimization
14. ICLR 2025Fast Summation of Radial Kernels via QMC Slicing
15. ICLR 2025Global Convergence in Neural ODEs: Impact of Activation Functions
16. ICLR 2025Improved Algorithms for Kernel Matrix-Vector Multiplication Under Sparsity Assumptions
17. ICLR 2025Learning Neural Networks with Distribution Shift: Efficiently Certifiable Guarantees
18. ICML 2025Adaptive kernel predictors from feature-learning infinite limits of neural networks
19. ICML 2025Beyond the Permutation Symmetry of Transformers: The Role of Rotation for Model Fusion
20. ICML 2025COKE: Core Kernel for More Efficient Approximation of Kernel Weights in Multiple Kernel Clustering
21. ICML 2025Equivariant Neural Tangent Kernels
22. ICML 2025K$^2$IE: Kernel Method-based Kernel Intensity Estimators for Inhomogeneous Poisson Processes
23. ICML 2025Learning Curves of Stochastic Gradient Descent in Kernel Regression
24. ICML 2025Learning Input Encodings for Kernel-Optimal Implicit Neural Representations
25. ICML 2025Nested Expectations with Kernel Quadrature
26. ICML 2025NTK-DFL: Enhancing Decentralized Federated Learning in Heterogeneous Settings via Neural Tangent Kernel
27. ICML 2025PEAKS: Selecting Key Training Examples Incrementally via Prediction Error Anchored by Kernel Similarity
28. ICML 2025Permutation-Free High-Order Interaction Tests
29. ICML 2025RollingQ: Reviving the Cooperation Dynamics in Multimodal Transformer
30. ICML 2025Scalable Gaussian Processes with Latent Kronecker Structure
31. ICML 2025The Price of Linear Time: Error Analysis of Structured Kernel Interpolation
32. NeurIPS 2025Anchor-based Maximum Discrepancy for Relative Similarity Testing
33. NeurIPS 2025Beyond Benign Overfitting in Nadaraya-Watson Interpolators
34. NeurIPS 2025ClusterFusion: Expanding Operator Fusion Scope for LLM Inference via Cluster-Level Collective Primitive
35. NeurIPS 2025Computational Efficiency under Covariate Shift in Kernel Ridge Regression
36. NeurIPS 2025Conditional Distribution Compression via the Kernel Conditional Mean Embedding
37. NeurIPS 2025Diversifying Parallel Ergodic Search: A Signature Kernel Evolution Strategy
38. NeurIPS 2025Fast Training of Large Kernel Models with Delayed Projections
39. NeurIPS 2025FFN Fusion: Rethinking Sequential Computation in Large Language Models
40. NeurIPS 2025FlashMoE: Fast Distributed MoE in a Single Kernel
41. NeurIPS 2025Hierarchical Shortest-Path Graph Kernel Network
42. NeurIPS 2025Just One Layer Norm Guarantees Stable Extrapolation
43. NeurIPS 2025Kernel conditional tests from learning-theoretic bounds
44. NeurIPS 2025Kernel von Mises Formula of the Influence Function
45. NeurIPS 2025Learning Provably Improves the Convergence of Gradient Descent
46. NeurIPS 2025Linearization Explains Fine-Tuning in Large Language Models
47. NeurIPS 2025MANGO: Multimodal Attention-based Normalizing Flow Approach to Fusion Learning
48. NeurIPS 2025Multi-Kernel Correlation-Attention Vision Transformer for Enhanced Contextual Understanding and Multi-Scale Integration
49. NeurIPS 2025On the Hardness of Conditional Independence Testing In Practice
50. NeurIPS 2025Optimal kernel regression bounds under energy-bounded noise

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。