Aspects of scaling and scalability for flow-based sampling of lattice QCD
Abstract Recent applications of machine-learned normalizing flows to sampling in lattice field theory suggest that such methods may be able to mitigate critical slowing down and topological freezing. However, these demonstrations have been at the scale of toy models, and it remains t...
| Main Authors: | , , , , , , , , , , , , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | English |
| Published: |
Springer Berlin Heidelberg
2023
|
| Online Access: | https://hdl.handle.net/1721.1/152908 |
| _version_ | 1826202557054189568 |
|---|---|
| author | Abbott, Ryan Albergo, Michael S. Botev, Aleksandar Boyda, Denis Cranmer, Kyle Hackett, Daniel C. Matthews, Alexander G. D. G. Racanière, Sébastien Razavi, Ali Rezende, Danilo J. Romero-López, Fernando Shanahan, Phiala E. Urban, Julian M. |
| author2 | Massachusetts Institute of Technology. Center for Theoretical Physics |
| author_facet | Massachusetts Institute of Technology. Center for Theoretical Physics Abbott, Ryan Albergo, Michael S. Botev, Aleksandar Boyda, Denis Cranmer, Kyle Hackett, Daniel C. Matthews, Alexander G. D. G. Racanière, Sébastien Razavi, Ali Rezende, Danilo J. Romero-López, Fernando Shanahan, Phiala E. Urban, Julian M. |
| author_sort | Abbott, Ryan |
| collection | MIT |
| description | Abstract
Recent applications of machine-learned normalizing flows to sampling in lattice field theory suggest that such methods may be able to mitigate critical slowing down and topological freezing. However, these demonstrations have been at the scale of toy models, and it remains to be determined whether they can be applied to state-of-the-art lattice quantum chromodynamics calculations. Assessing the viability of sampling algorithms for lattice field theory at scale has traditionally been accomplished using simple cost scaling laws, but as we discuss in this work, their utility is limited for flow-based approaches. We conclude that flow-based approaches to sampling are better thought of as a broad family of algorithms with different scaling properties, and that scalability must be assessed experimentally. |
| first_indexed | 2024-09-23T12:09:22Z |
| format | Article |
| id | mit-1721.1/152908 |
| institution | Massachusetts Institute of Technology |
| language | English |
| last_indexed | 2024-09-23T12:09:22Z |
| publishDate | 2023 |
| publisher | Springer Berlin Heidelberg |
| record_format | dspace |
| spelling | mit-1721.1/1529082024-01-24T18:47:39Z Aspects of scaling and scalability for flow-based sampling of lattice QCD Abbott, Ryan Albergo, Michael S. Botev, Aleksandar Boyda, Denis Cranmer, Kyle Hackett, Daniel C. Matthews, Alexander G. D. G. Racanière, Sébastien Razavi, Ali Rezende, Danilo J. Romero-López, Fernando Shanahan, Phiala E. Urban, Julian M. Massachusetts Institute of Technology. Center for Theoretical Physics Abstract Recent applications of machine-learned normalizing flows to sampling in lattice field theory suggest that such methods may be able to mitigate critical slowing down and topological freezing. However, these demonstrations have been at the scale of toy models, and it remains to be determined whether they can be applied to state-of-the-art lattice quantum chromodynamics calculations. Assessing the viability of sampling algorithms for lattice field theory at scale has traditionally been accomplished using simple cost scaling laws, but as we discuss in this work, their utility is limited for flow-based approaches. We conclude that flow-based approaches to sampling are better thought of as a broad family of algorithms with different scaling properties, and that scalability must be assessed experimentally. 2023-11-06T16:29:24Z 2023-11-06T16:29:24Z 2023-11-04 2023-11-05T04:12:06Z Article http://purl.org/eprint/type/JournalArticle https://hdl.handle.net/1721.1/152908 The European Physical Journal A. 2023 Nov 04;59(11):257 PUBLISHER_CC en https://doi.org/10.1140/epja/s10050-023-01154-w Creative Commons Attribution https://creativecommons.org/licenses/by/4.0/ The Author(s) application/pdf Springer Berlin Heidelberg Springer Berlin Heidelberg |
| spellingShingle | Abbott, Ryan Albergo, Michael S. Botev, Aleksandar Boyda, Denis Cranmer, Kyle Hackett, Daniel C. Matthews, Alexander G. D. G. Racanière, Sébastien Razavi, Ali Rezende, Danilo J. Romero-López, Fernando Shanahan, Phiala E. Urban, Julian M. Aspects of scaling and scalability for flow-based sampling of lattice QCD |
| title | Aspects of scaling and scalability for flow-based sampling of lattice QCD |
| title_full | Aspects of scaling and scalability for flow-based sampling of lattice QCD |
| title_fullStr | Aspects of scaling and scalability for flow-based sampling of lattice QCD |
| title_full_unstemmed | Aspects of scaling and scalability for flow-based sampling of lattice QCD |
| title_short | Aspects of scaling and scalability for flow-based sampling of lattice QCD |
| title_sort | aspects of scaling and scalability for flow based sampling of lattice qcd |
| url | https://hdl.handle.net/1721.1/152908 |
| work_keys_str_mv | AT abbottryan aspectsofscalingandscalabilityforflowbasedsamplingoflatticeqcd AT albergomichaels aspectsofscalingandscalabilityforflowbasedsamplingoflatticeqcd AT botevaleksandar aspectsofscalingandscalabilityforflowbasedsamplingoflatticeqcd AT boydadenis aspectsofscalingandscalabilityforflowbasedsamplingoflatticeqcd AT cranmerkyle aspectsofscalingandscalabilityforflowbasedsamplingoflatticeqcd AT hackettdanielc aspectsofscalingandscalabilityforflowbasedsamplingoflatticeqcd AT matthewsalexandergdg aspectsofscalingandscalabilityforflowbasedsamplingoflatticeqcd AT racanieresebastien aspectsofscalingandscalabilityforflowbasedsamplingoflatticeqcd AT razaviali aspectsofscalingandscalabilityforflowbasedsamplingoflatticeqcd AT rezendedaniloj aspectsofscalingandscalabilityforflowbasedsamplingoflatticeqcd AT romerolopezfernando aspectsofscalingandscalabilityforflowbasedsamplingoflatticeqcd AT shanahanphialae aspectsofscalingandscalabilityforflowbasedsamplingoflatticeqcd AT urbanjulianm aspectsofscalingandscalabilityforflowbasedsamplingoflatticeqcd |