Multimodal Transformer Based Generic Mixture Density Network for Scattering Timescale Estimation of Fast Radio Bursts

2026-06-02

AI Summary

This study addresses the inefficiency, manual dependency, and sensitivity to initial conditions of conventional methods for measuring the scattering timescale (τ) of fast radio bursts (FRBs), which struggle to meet the demands of high discovery rates. To overcome these limitations, the authors propose the Multimodal Transformer Gaussian Mixture Density Network (MT-GMDN), which uniquely integrates a multimodal Transformer with a mixture density network to jointly encode FRB dynamic spectra and temporal profiles. This framework probabilistically models the zero-inflated distribution of τ, effectively accommodating both measurable scattering events and non-scattering samples. Evaluated on CHIME/FRB data, the method achieves an R² of 94% and a recall rate of 90%, while simultaneously providing point estimates and well-calibrated uncertainty intervals for τ.

Abstract

The discovery rate of fast radio bursts (FRBs) continues to increase with the advent of new radio facilities and yet extracting their astrophysical parameters such as scattering timescale ($τ$) remains a significant bottleneck. Current $τ$ measurement approaches like fitting analytic template models and scattering aware de-convolution are accurate but slow, sensitive to initialization, limited by low signal to noise and often require manual supervision. These limitations inspired us to explore fast, robust and scalable machine learning methods to estimate the astrophysical parameter value. We present a deep learning approach named Multimodal Transformer Based Generic Mixture Density Network (MT-GMDN) which ingests FRB dynamic spectrum and its corresponding timeseries profile through parallel transformer encoders, fuses their latent representations and predicts the distribution of $τ$ with probabilistic output derived from generic mixture-density formulation. This formulation not only estimates the value of $τ$ but also captures the (zero inflated) nature of FRB populations where a significant fraction of bursts exhibit unresolvable scattering. We trained MT-GMDN on $\sim3500$ FRBs from CHIME/FRB \cattwo while holding out some fraction of FRBs for validation during training and for testing after the training completes. The model achieves a coefficient of determination ($R^2$) value of $94\%$ on the expected value of $τ$ for the events with measurable scattering with an excellent recall value of $90\%$ on the test data set. The model was also able to incorporate heteroskedastic errors enabling us the construction of a confidence interval for the predictions.

Problem

• Fast Radio Bursts
• Scattering Timescale
• Astrophysical Parameter Estimation
• Signal-to-Noise Ratio
• Parameter Extraction

Innovation

• Multimodal Transformer
• Mixture Density Network
• Scattering Timescale Estimation
• Fast Radio Bursts
• Probabilistic Deep Learning