A taxonomy of automatic differentiation pitfalls - Inria - Institut national de recherche en sciences et technologies du numérique
Article Dans Une Revue Wiley Interdisciplinary Reviews: Data Mining and Knowledge Discovery Année : 2024

A taxonomy of automatic differentiation pitfalls

Résumé

Abstract Automatic differentiation is a popular technique for computing derivatives of computer programs. While automatic differentiation has been successfully used in countless engineering, science, and machine learning applications, it can sometimes nevertheless produce surprising results. In this paper, we categorize problematic usages of automatic differentiation, and illustrate each category with examples such as chaos, time‐averages, discretizations, fixed‐point loops, lookup tables, linear solvers, and probabilistic programs, in the hope that readers may more easily avoid or detect such pitfalls. We also review debugging techniques and their effectiveness in these situations. This article is categorized under: Technologies > Machine Learning

Dates et versions

hal-04693411 , version 1 (10-09-2024)

Licence

Identifiants

Citer

Jan Hückelheim, Harshitha Menon, William Moses, Bruce Christianson, Paul Hovland, et al.. A taxonomy of automatic differentiation pitfalls. Wiley Interdisciplinary Reviews: Data Mining and Knowledge Discovery, 2024, ⟨10.1002/widm.1555⟩. ⟨hal-04693411⟩
14 Consultations
0 Téléchargements

Altmetric

Partager

More