Differentiable Cosmological Simulation with Adjoint Method
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2022
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866917587815759872 |
|---|---|
| author | Li, Yin Modi, Chirag Jamieson, Drew Zhang, Yucheng Lu, Libin Feng, Yu Lanusse, François Greengard, Leslie |
| author_facet | Li, Yin Modi, Chirag Jamieson, Drew Zhang, Yucheng Lu, Libin Feng, Yu Lanusse, François Greengard, Leslie |
| contents | Rapid advances in deep learning have brought not only myriad powerful neural networks, but also breakthroughs that benefit established scientific research. In particular, automatic differentiation (AD) tools and computational accelerators like GPUs have facilitated forward modeling of the Universe with differentiable simulations. Based on analytic or automatic backpropagation, current differentiable cosmological simulations are limited by memory, and thus are subject to a trade-off between time and space/mass resolution, usually sacrificing both. We present a new approach free of such constraints, using the adjoint method and reverse time integration. It enables larger and more accurate forward modeling at the field level, and will improve gradient based optimization and inference. We implement it in an open-source particle-mesh (PM) $N$-body library pmwd (particle-mesh with derivatives). Based on the powerful AD system JAX, pmwd is fully differentiable, and is highly performant on GPUs. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2211_09815 |
| institution | arXiv |
| publishDate | 2022 |
| record_format | arxiv |
| spellingShingle | Differentiable Cosmological Simulation with Adjoint Method Li, Yin Modi, Chirag Jamieson, Drew Zhang, Yucheng Lu, Libin Feng, Yu Lanusse, François Greengard, Leslie Instrumentation and Methods for Astrophysics Cosmology and Nongalactic Astrophysics Rapid advances in deep learning have brought not only myriad powerful neural networks, but also breakthroughs that benefit established scientific research. In particular, automatic differentiation (AD) tools and computational accelerators like GPUs have facilitated forward modeling of the Universe with differentiable simulations. Based on analytic or automatic backpropagation, current differentiable cosmological simulations are limited by memory, and thus are subject to a trade-off between time and space/mass resolution, usually sacrificing both. We present a new approach free of such constraints, using the adjoint method and reverse time integration. It enables larger and more accurate forward modeling at the field level, and will improve gradient based optimization and inference. We implement it in an open-source particle-mesh (PM) $N$-body library pmwd (particle-mesh with derivatives). Based on the powerful AD system JAX, pmwd is fully differentiable, and is highly performant on GPUs. |
| title | Differentiable Cosmological Simulation with Adjoint Method |
| topic | Instrumentation and Methods for Astrophysics Cosmology and Nongalactic Astrophysics |
| url | https://arxiv.org/abs/2211.09815 |