Computational Graphs and Rounding Error

By F. L. Bauer. 1974

How researchers see this paper

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

1. Yoshua Bengio
2. Sergey Levine
3. Tengyu Ma
4. Ruslan Salakhutdinov
5. Zhe Gan
6. Yoav Goldberg
7. Mihaela Rosca
8. Quoc V. Le

Citations in the corpus

• A Survey on Neural Architecture Search (2019, background): “This can be formalized in the language of computational graphs (Bauer, 1974) where neural networks are represented as directed acyclic graphs with a set of nodes Z.”
• Automatic differentiation in machine learning: a survey (2015, methodology): “A given trace of elementary operations can also be represented using a computational graph (Bauer, 1974), as shown in Fig.”
• Monte Carlo Gradient Estimation in Machine Learning (2019, background): “on of the stochastic gradient estimators described in this work, see the note on the code in the introduction. 9.2 Stochastic Computational Graphs and Probabilistic Programming. Computational graphs (Bauer, 1974) are now the standard formalism with which we represent complex mathematical systems in ”
• Data-dependent Sample Complexity of Deep Neural Networks via Lipschitz Augmentation (2019, background): “Computational graphs are originally introduced by Bauer (1974) to represent computational processes and study error propagation.”