Corpus ID: 233210711

Getting to the Point. Index Sets and Parallelism-Preserving Autodiff for Pointful Array Programming

@article{Paszke2021GettingTT,
  title={Getting to the Point. Index Sets and Parallelism-Preserving Autodiff for Pointful Array Programming},
  author={Adam Paszke and Daniel Johnson and D. Duvenaud and Dimitrios Vytiniotis and Alexey Radul and Matthew Johnson and Jonathan Ragan-Kelley and D. Maclaurin},
  journal={ArXiv},
  year={2021},
  volume={abs/2104.05372}
}
We present a novel programming language design that attempts to combine the clarity and safety of high-level functional languages with the efficiency and parallelism of low-level numerical languages. We treat arrays as eagerly-memoized functions on typed index sets, allowing abstract function manipulations, such as currying, to work on arrays. In contrast to composing primitive bulk-array operations, we argue for an explicit nested indexing style that mirrors application of functions to… Expand
1 Citations

Figures from this paper

Decomposing reverse-mode automatic differentiation

References

SHOWING 1-10 OF 58 REFERENCES
Secure distributed programming with value-dependent types
TensorFlow: A system for large-scale machine learning
Harnessing the Multicores: Nested Data Parallelism in Haskell
Efficient Differentiable Programming
  • Machine Learning and Programming
  • 2019
Definition of standard ML
Automatic differentiation in machine learning: a survey
AD on Higher Order Functions
  • 2021
Array programming with NumPy
Compiling generalized histograms for GPU
DiffTaichi: Differentiable Programming for Physical Simulation
...
1
2
3
4
5
...