mlsanityData to 7 Oct 2026

Kernel Operations on the GPU, with Autodiff, without Memory Overflows

arXiv 2004.11127 ↗First posted 27 Mar 2020cs.LGcited by 101

Not on the map. Why?

Abstract (arXiv)

The KeOps library provides a fast and memory-efficient GPU support for tensors whose entries are given by a mathematical formula, such as kernel and distance matrices. KeOps alleviates the major bottleneck of tensor-centric libraries for kernel and geometric applications: memory consumption. It also supports automatic differentiation and outperforms standard GPU baselines, including PyTorch CUDA tensors or the Halide and TVM libraries. KeOps combines optimized C++/CUDA schemes with binders for high-level languages: Python (Numpy and PyTorch), Matlab and GNU R. As a result, high-level "quadratic" codes can now scale up to large data sets with millions of samples processed in seconds. KeOps brings graphics-like performances for kernel methods and is freely available on standard repositories (PyPi, CRAN). To showcase its versatility, we provide tutorials in a wide range of settings online at urlwww.kernel-operations.io.

Key facts. Cited by 101 papers in the map (data through 7 Oct 2026). First posted on arXiv 27 Mar 2020.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 20; 2022: 25; 2023: 14; 2024: 13; 2025: 16; 2026: 13.

101 papers.

Sort:

By area:

  1. 2022FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awarenesscited by 3,611
  2. 2022Diagonal State Spaces are as Effective as Structured State Spacescited by 350
  3. 2021Sinkformers: Transformers with Doubly Stochastic Attentioncited by 136
  4. 2022Liquid Structural State-Space Modelscited by 134
  5. 2021Shape registration in the time of transformerscited by 51
  6. 2021Accurate Point Cloud Registration with Robust Optimal Transportcited by 38
  7. 2021On UMAP's true loss functioncited by 32
  8. 2021Towards a mathematical theory of trajectory inferencecited by 31
  9. 2022From t-SNE to UMAP with contrastive learningcited by 28
  10. 2022Posterior and Computational Uncertainty in Gaussian Processescited by 23
  11. 2022Latent Graph Inference using Product Manifoldscited by 23
  12. 2022Simplifying and Understanding State Space Models with Diagonal Linear RNNscited by 23
38 more
  1. 2021Preconditioning for Scalable Gaussian Process Hyperparameter Optimizationcited by 22
  2. 2021Tighter Bounds on the Log Marginal Likelihood of Gaussian Process Regression Using Conjugate Gradientscited by 20
  3. 2022Elastic shape analysis of surfaces with second-order Sobolev metrics: a comprehensive numerical frameworkcited by 18
  4. 2021Decoupled Exploration and Exploitation Policies for Sample-Efficient Reinforcement Learningcited by 14
  5. 2024Memory-Scalable and Simplified Functional Map Learningcited by 14
  6. 2021ResNet-LDDMM: Advancing the LDDMM Framework using Deep Residual Networkscited by 13
  7. 2021Fast and Scalable Optimal Transport for Brain Tractogramscited by 11
  8. 2022Efficient Hyperparameter Tuning for Large Scale Kernel Ridge Regressioncited by 11
  9. 2022Fast Nonlinear Vector Quantile Regressioncited by 11
  10. 2023qPOTS: Efficient batch multiobjective Bayesian optimization via Pareto optimal Thompson samplingcited by 11
  11. 2023Large-Scale Gaussian Processes via Alternating Projectioncited by 11
  12. 2022Low-Precision Arithmetic for Fast Gaussian Processescited by 10
  13. 2021SKIing on Simplices: Kernel Interpolation on the Permutohedral Lattice for Scalable Gaussian Processescited by 9
  14. 2021Towards Efficient Point Cloud Graph Neural Networks Through Architectural Simplificationcited by 9
  15. 2024Conditional simulation via entropic optimal transport: Toward non-parametric estimation of conditional Brenier mapscited by 9
  16. 2025Position: Curvature Matrices Should Be Democratized via Linear Operatorscited by 9
  17. 2025ELECTRA: A Cartesian Network for 3D Charge Density Prediction with Floating Orbitalscited by 9
  18. 2022Single-Photon Structured Lightcited by 8
  19. 20223D Shape Sequence of Human Comparison and Classification using Current and Varifoldscited by 8
  20. 2023CoLA: Exploiting Compositional Structure for Automatic and Efficient Numerical Linear Algebracited by 7
  21. 2024Advancing Deep Active Learning & Data Subset Selection: Unifying Principles with Information-Theory Intuitionscited by 7
  22. 2024Fast Kernel Summation in High Dimensions via Slicing and Fourier Transformscited by 7
  23. 2024Fast Summation of Radial Kernels via QMC Slicingcited by 7
  24. 2026FlashSinkhorn: IO-Aware Entropic Optimal Transport on GPUcited by 7
  25. 2022Unsupervised Voice-Face Representation Learning by Cross-Modal Prototype Contrastcited by 6
  26. 2023Toward Mesh-Invariant 3D Generative Deep Learning with Geometric Measurescited by 6
  27. 2023Efficient and Accurate Optimal Transport with Mirror Descent and Conjugate Gradientscited by 6
  28. 2023Neural Snowflakes: Universal Latent Graph Inference via Trainable Latent Geometriescited by 6
  29. 2022Self-Supervised Correspondence Estimation via Multiview Registrationcited by 5
  30. 2023Label Shift Quantification with Robustness Guarantees via Distribution Feature Matchingcited by 5
  31. 2025A Truncated Newton Method for Optimal Transportcited by 5
  32. 2021Unrolled Variational Bayesian Algorithm for Image Blind Deconvolutioncited by 4
  33. 2021Partial Matching in the Space of Varifoldscited by 3
  34. 2021Low-rank statistical finite elements for scalable model-data synthesiscited by 3
  35. 2021Nonlinearities in Steerable SO(2)-Equivariant CNNscited by 3
  36. 2022Neural Bregman Divergences for Distance Learningcited by 3
  37. 2022Memory Safe Computations with XLA Compilercited by 3
  38. 2022GNPM: Geometric-Aware Neural Parametric Modelscited by 3

50 most cited of 101 All 101 (JSON)