NASA Cosmic Origins · AI/ML Science and Technology Interest Group

Deep Learning for Astrophysics

A hands-on textbook on modern artificial intelligence for astronomy and astrophysics — from autodifferentiation and neural architectures to generative models, simulation-based inference, and autonomous research agents.

23
Chapters
6
Parts
17
Lecturers
Runnable
Code & Outputs
About this book

This Textbook Edition was curated from the lecture series of the NASA Cosmic Origins AI/ML Science and Technology Interest Group (STIG). We homogenized the original lectures, notebooks, and supporting materials into a consistent chapter format while preserving the executable notebooks and their real outputs.

For the original schedule, recordings, notebooks, slides, and source lecture materials, see the NASA Cosmic Origins AI/ML STIG site.

The arc runs from computational foundations through the full zoo of deep-learning architectures, into generative modeling and inference, reinforcement learning, and the large-language-model agents now reshaping the research workflow — closing with the broader questions of instrumentation, publishing, and scientific understanding.

Table of Contents

Contributors

Chapter Authors

  • Yuan-Sen Ting3 lectures
  • Phill Cargile2 lectures
  • Carol Cuesta-Lazaro2 lectures
  • Tomasz Rozanski2 lectures
  • Anna Scaife2 lectures
  • John F. Wu2 lectures
  • André Curtis-Trudel1 lecture
  • Gregory Green1 lecture
  • Ryan McClelland1 lecture
  • Daniel Muthukrishna1 lecture
  • Tri Nguyen1 lecture
  • Helen Qu1 lecture
  • Jesse Thaler1 lecture
  • Licia Verde1 lecture
  • Francisco Villaescusa-Navarro1 lecture
  • Duo Xu1 lecture
  • Siyu Yao1 lecture

Leadership Council

  • Yuan-Sen Ting (Co-Chair)The Ohio State University
  • Digvijay Wadekar (Co-Chair)University of Texas at Austin
  • Andrew SaydjariPrinceton University
  • Alex GaglianoMIT
  • Carol Cuesta-LazaroInstitute for Advanced Study at Princeton / Flatiron Institute
  • Georgios ValogiannisUniversity of Chicago
  • Siddharth Mishra-SharmaBoston University