Deep learning powered numerical relativity surrogate for binary black hole waveforms
Gravitational-wave (GW) approximants are essential for gravitational-wave astronomy, allowing the coverage of the binary black hole parameter space for inference or match filtering without costly numerical relativity (NR) simulations but generally trading some accuracy for computational efficiency. To reduce this trade-off, NR surrogate models can be constructed using interpolation within NR waveform space. We present a two-stage training approach for neural network-based NR surrogate models. We initially train four models on waveforms generated from four different GW approximants and then fine-tune these models on NR data. We show that despite the median mismatches of the pretrained models with NR ranging over two orders of magnitude, the fine-tuned models all reach median mismatches of order ${10}^{\ensuremath{-}5}$, on par with top-performing NR surrogates. The dual-stage artificial neural surrogate (DANSur3dq8) models also offer rapid waveform generation, with millions of waveforms being generated in under 20 ms on a GPU. Implemented in the bilby framework, we show DANSur3dq8 can be used for parameter estimation tasks.
Authors
- O. Freitas (ORCID: https://orcid.org/0000-0002-2898-1256)
- A. Onofre (ORCID: https://orcid.org/0000-0003-3471-2703)
- A. Torres-Forné (ORCID: https://orcid.org/0000-0001-8709-5118)
- A. Theodoropoulos (ORCID: https://orcid.org/0000-0003-4486-7135)
- Solange Nunes (ORCID: https://orcid.org/0000-0002-1506-3368)
- José D. Martín‐Guerrero (ORCID: https://orcid.org/0000-0001-9378-0285)
- José A. Font (ORCID: https://orcid.org/0000-0001-6650-2634)
- Tiago Fernandes (ORCID: https://orcid.org/0009-0006-6820-2065)
- Nino Villanueva
Institutions
- Universitat de València (ES)
- Artificial Intelligence Research Institute (ES)
- Research Network (United States) (US)
- University of Minho (PT)
Publication Details
- Journal
- Physical review. D/Physical review. D.
- Published
- 2025-08-25
- DOI
- https://doi.org/10.1103/7bkx-hs53
- Citations
- 3
- Primary Topic
- Pulsars and Gravitational Waves Research
- Type
- article
- Field-Weighted Citation Impact
- 3.46
Funders
- National Science Foundation
- European Commission
- Ministry of Education, Culture, Sports, Science and Technology
- Academia Sinica
- Generalitat Valenciana
- National Research Foundation of Korea
- Max-Planck-Gesellschaft
- Centre National de la Recherche Scientifique
- Ministerio de Ciencia e Innovación
- Ministry of Science and ICT, South Korea
- Science and Technology Facilities Council
- Australian Research Council
- Japan Society for the Promotion of Science
- Fundação para a Ciência e a Tecnologia
- Consejo Superior de Investigaciones Científicas
- Instituto Nazionale di Fisica Nucleare
- European Regional Development Fund
- Agencia Estatal de Investigación
- Conselleria de Innovación, Universidades, Ciencia y Sociedad Digital, Generalitat Valenciana
- HORIZON EUROPE Marie Sklodowska-Curie Actions