Reverse-mode AD in a functional framework: Lambda the ultimate backpropagator

By Barak A. Pearlmutter, Jeffrey Mark Siskind. 2008

116 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computational Theory and Mathematics, Hardware and Architecture

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Quoc V. Le
2. Yoshua Bengio
3. Sergey Levine
4. Mohit Bansal
5. Ruslan Salakhutdinov
6. Pieter Abbeel
7. Martín Abadi
8. Tim Harley

In Computational Theory and Mathematics

1. Frédo Durand
2. Manjunath Kudlur
3. Martín Abadi
4. Andy Davis
5. Joseph Dean

In Hardware and Architecture

1. Derek Gordon Murray
2. Michael Isard
3. Martín Abadi
4. Jonathan Ragan-Kelley
5. Paul Barham
6. Frédo Durand
7. Sanjay Ghemawat
8. M. Burrows

Citations in the corpus

• DiffTaichi: Differentiable Programming for Physical Simulation (2019, methodology): “Earlier work on automatic differentiation focuses on transforming existing scalar code to obtain derivatives (e.g. Utke et al. (2008), Hascoet & Pascual (2013), Pearlmutter & Siskind (2008)).”
• Dynamic control flow in large-scale machine learning (2018, background): “gorithms and also the application of reinforcement learning to swapping and recomputation decisions. Finally, our research is related to a substantial body of work on automatic differentiation (e.g., [18, 25, 32]). That work includes systems for automatic differentiation of “ordinary” programming la”