[2403.14606] The Elements of Differentiable Programming
Abstract:Artificial intelligence has recently experienced remarkable advances, fueled by large models, vast datasets, accelerated hardware, and, last but not least, the transformative power of differentiable programming. This new programming paradigm enables end-to-end differentiation of complex computer programs (including those with control flows and data structures), making gradient-based optimization of program parameters possible. As an emerging paradigm, differentiable programming builds upon several areas of computer science and applied mathematics, including automatic differentiation, graphical models, optimization and statistics. This book presents a comprehensive review of the fundamental concepts useful for differentiable programming. We adopt two main perspectives, that of optimization and that of probability, with clear analogies between the two. Differentiable programming is not merely the differentiation of programs, but also the thoughtful design of programs intended for differentiation. By making programs differentiable, we inherently introduce probability distributions over their execution, providing a means to quantify the uncertainty associated with program outputs.
The Elements of Differentiable Programming Mathieu Blondel Google DeepMind mblondel@google.com arXiv:2403.14606v4 [cs.LG] 3 Aug 2026 Vincent Roulet…
saved by
related reading
- course_stat_rl.pdfmit.edu
- nn-notes.pdfboris-hanin.github.io
- 2404.17625arxiv.org
- The Little Book of Deep Learningfleuret.org
- Differentiable Programming from Scratchthenumb.at
- [2604.21691] There Will Be a Scientific Theory of Deep Learningarxiv.org
- Pen and Paper Exercises in Machine Learningarxiv.org
- Understanding Deep Learningudlbook.github.io
- [2604.21691] There Will Be a Scientific Theory of Deep Learningarxiv.org
- [2604.07242] Weaves, Wires, and Morphisms: Formalizing and Implementing the Algebra of Deep Learningarxiv.org
- Book: Alice’s Adventures in a differentiable wonderlandsscardapane.it
- [2605.01172] A Theory of Generalization in Deep Learningarxiv.org