Backpropagation in the simply typed lambda-calculus with linear negation

Author:

Brunel Aloïs1,Mazza Damiano2,Pagani Michele3

Affiliation:

1. Deepomatic, France

2. CNRS, France

3. IRIF, France / University of Paris Diderot, France

Abstract

Backpropagation is a classic automatic differentiation algorithm computing the gradient of functions specified by a certain class of simple, first-order programs, called computational graphs. It is a fundamental tool in several fields, most notably machine learning, where it is the key for efficiently training (deep) neural networks. Recent years have witnessed the quick growth of a research field called differentiable programming, the aim of which is to express computational graphs more synthetically and modularly by resorting to actual programming languages endowed with control flow operators and higher-order combinators, such as map and fold. In this paper, we extend the backpropagation algorithm to a paradigmatic example of such a programming language: we define a compositional program transformation from the simply-typed lambda-calculus to itself augmented with a notion of linear negation, and prove that this computes the gradient of the source program with the same efficiency as first-order backpropagation. The transformation is completely effect-free and thus provides a purely logical understanding of the dynamics of backpropagation.

Publisher

Association for Computing Machinery (ACM)

Subject

Safety, Risk, Reliability and Quality,Software

Reference30 articles.

1. An Abstract Factorization Theorem for Explicit Substitutions;Accattoli Beniamino;Proceedings of RTA (LIPIcs),2012

2. Proof Nets and the Linear Substitution Calculus

3. Distilling abstract machines

4. Environments and the complexity of abstract machines

Cited by 26 articles. 订阅此论文施引文献 订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献

1. Verifying an Effect-Handler-Based Define-By-Run Reverse-Mode AD Library;Logical Methods in Computer Science;2023-10-23

2. Taylor Expansion as a Monad in Models of DiLL;2023 38th Annual ACM/IEEE Symposium on Logic in Computer Science (LICS);2023-06-26

3. ωPAP Spaces: Reasoning Denotationally About Higher-Order, Recursive Probabilistic and Differentiable Programs;2023 38th Annual ACM/IEEE Symposium on Logic in Computer Science (LICS);2023-06-26

4. Computer-aided diagnosis systems: a comparative study of classical machine learning versus deep learning-based approaches;Knowledge and Information Systems;2023-05-24

5. Smart Choices and the Selection Monad;Logical Methods in Computer Science;2023-04-20

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3