mlsanityData to 14 Oct 2026

Software-based Automatic Differentiation is Flawed

arXiv 2305.03863 ↗First posted 5 May 2023cs.LGcited by 3

Not on the map. Why?

Abstract (arXiv)

Various software efforts embrace the idea that object oriented programming enables a convenient implementation of the chain rule, facilitating so-called automatic differentiation via backpropagation. Such frameworks have no mechanism for simplifying the expressions (obtained via the chain rule) before evaluating them. As we illustrate below, the resulting errors tend to be unbounded.

Key facts. Cited by 3 papers in the map (data through 14 Oct 2026). Builds on 5 of its 15 references that are in the map. First posted on arXiv 5 May 2023.

Builds on

5 of 15 references.

References · 5

  1. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,305
  2. 2016TensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systemscited by 664
  3. 2014Caffe: Convolutional Architecture for Fast Feature Embeddingcited by 288
  4. 2016Theano: A Python framework for fast computation of mathematical expressionscited by 64
  5. 2022Analytically Integratable Zero-restlength Springs for Capturing Dynamic Modes unrepresented by Quasistatic Neural Networkscited by 2

Built on by

Cited by year

Citing papers per year: 2023: 3.

3 papers.

  1. 2023Neural-Integrated Meshfree (NIM) Method: A differentiable programming-based hybrid solver for computational mechanicscited by 4
  2. 2023Addressing Discontinuous Root-Finding for Subsequent Differentiability in Machine Learning, Inverse Problems, and Controlcited by 0
  3. 2023Zero Coordinate Shift: Whetted Automatic Differentiation for Physics-informed Operator Learningcited by 0