Physics-Informed Neural Networks and Data-Driven Models for GRB X-ray Light-Curve Gap Reconstruction
Abstract
Temporal gaps in Swift-XRT X-ray afterglows of gamma-ray bursts (GRBs) limit how precisely the standard empirical plateau function constrains three parameters: the plateau end time Ta, the plateau flux Fa, and the post-plateau decay index α. These parameters underpin the luminosity-time and fundamental-plane correlations of GRB plateaus so that smaller measurement uncertainties can improve the statistical power of GRBs as cosmological probes. As the fifth study in a series on light-curve reconstruction, this work benchmarks four models on a parent sample of 545 Swift-XRT GRBs: a Physics-Informed Neural Network (PINN) with seven afterglow-motivated priors, ReFANN, a Siamese dual-branch network with Monte Carlo dropout, and Polynomial Quantile Regression (PQR). All four reduce the fractional uncertainties in log Ta, log Fa, and α relative to the original observations. The PINN with a broken power-law prior gives the largest reductions (∼41-49\%), at a higher outlier rate (∼15-20\%), while ReFANN, Siamese network and PQR give more modest reductions (∼19-27\%) with outlier fractions ≲4\%, across a wider range of light-curve morphologies. A reduced-χ2 prior-selection scheme recovers a morphological classification consistent with independent labelling, though it does not match the best single-prior reduction. Together, these results provide a systematic benchmark of physics-informed and data-driven reconstruction strategies for irregularly sampled GRB afterglows.
Related
- Cosmo-PINN: A Physics-Informed Neural Network for Cosmological Reconstruction
- Reconstructing the generalized Barrow holographic dark energy with physics-informed neural networks
- Boosted Physics-Informed Neural Networks for Solving Homogeneous and Non-Homogeneous Lane-Emden Type Equations
- C3NN-SBI: Learning hierarchies of N -point statistics from cosmological fields with physics-informed neural networks
- Physics-Informed Neural Networks vs U-Net: Application to thermal dust emission recovery in the microwave
BibTeX
@misc{garg2026physicsinformed,
title = {{Physics-Informed Neural Networks and Data-Driven Models for GRB X-ray Light-Curve Gap Reconstruction}},
author = {Ayush Garg and Ritik Kumar and Maria G. Dainotti and Vipul Sharma and Amit Shukla and Dieter H. Hartmann},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2609.15261},
eprint = {2609.15261},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.15261},
}
RIS
TY - UNPB
TI - Physics-Informed Neural Networks and Data-Driven Models for GRB X-ray Light-Curve Gap Reconstruction
AU - Ayush Garg
AU - Ritik Kumar
AU - Maria G. Dainotti
AU - Vipul Sharma
AU - Amit Shukla
AU - Dieter H. Hartmann
PY - 2026
DA - 2026/10/12
PB - arXiv
DO - 10.48550/arxiv.2609.15261
UR - https://arxiv.org/abs/2609.15261
AB - Temporal gaps in Swift-XRT X-ray afterglows of gamma-ray bursts (GRBs) limit how precisely the standard empirical plateau function constrains three parameters: the plateau end time T_a, the plateau flux F_a, and the post-plateau decay index α. These parameters underpin the luminosity-time and fundamental-plane correlations of GRB plateaus so that smaller measurement uncertainties can improve the statistical power of GRBs as cosmological probes. As the fifth study in a series on light-curve reconstruction, this work benchmarks four models on a parent sample of 545 Swift-XRT GRBs: a Physics-Informed Neural Network (PINN) with seven afterglow-motivated priors, ReFANN, a Siamese dual-branch network with Monte Carlo dropout, and Polynomial Quantile Regression (PQR). All four reduce the fractional uncertainties in log T_a, log F_a, and α relative to the original observations. The PINN with a broken power-law prior gives the largest reductions (∼41-49\%), at a higher outlier rate (∼15-20\%), while ReFANN, Siamese network and PQR give more modest reductions (∼19-27\%) with outlier fractions ≲4\%, across a wider range of light-curve morphologies. A reduced-χ^2 prior-selection scheme recovers a morphological classification consistent with independent labelling, though it does not match the best single-prior reduction. Together, these results provide a systematic benchmark of physics-informed and data-driven reconstruction strategies for irregularly sampled GRB afterglows.
ER -