English

labrador: A domain-optimized machine-learning tool for gravitational wave inference

General Relativity and Quantum Cosmology 2026-04-13 v1 High Energy Astrophysical Phenomena Instrumentation and Methods for Astrophysics

Abstract

Fast and reliable inference of gravitational-wave source parameters is crucial for analyzing large catalogs that are reaching the size of hundreds of detections, and for identifying short-lived electromagnetic counterparts. Neural posterior estimation has emerged as a powerful inference method, where the model is trained on simulated gravitational-wave data at considerable computational cost, but thereafter enables extremely fast and inexpensive inference at test time. Here, we extend this approach by incorporating domain-specific physical insights and methods in the model architecture. These include compressing the data by heterodyning against a reference waveform chosen via approximate likelihood maximization, removing parameter degeneracies through tailored coordinate systems, and eliminating known multimodalities by folding the parameter space. As a result, the network is approximately equivariant to changes in the source parameters, and achieves a reduced training cost and improved model interpretability. Our implementation, called labrador, can be trained end-to-end on a 1-day timescale on 102\sim 10^2 CPU cores and a V100 GPU, achieving a median importance-sampling efficiency of 1% on quadrupolar, aligned-spin signals in a broad mass range (chirp mass M1-50M\mathcal M \in 1\text{-}50\,\mathrm{M}_\odot, mass ratio q>0.1q > 0.1). labrador is the first neural inference code to achieve extensive coverage of long-duration signals with secondary masses m2<10Mm_2 < 10\,\mathrm{M}_\odot, rendered possible by its equivariance property. Among our novel contributions is a numerically stable procedure that enables neural posterior estimation when the simulation and inference priors differ.

Keywords

Cite

@article{arxiv.2604.08897,
  title  = {labrador: A domain-optimized machine-learning tool for gravitational wave inference},
  author = {Javier Roulet and Marco Crisostomi and Lucy M. Thomas and Katerina Chatziioannou},
  journal= {arXiv preprint arXiv:2604.08897},
  year   = {2026}
}

Comments

22 pages, 10 figures, 2 tables

R2 v1 2026-07-01T12:02:18.028Z