Speaker
Description
Recent detections of very high-energy (VHE; $E > 100$\,GeV) emission from gamma-ray burst (GRB) afterglows have highlighted the need for accurate synchrotron self-Compton (SSC) models. Although numerical approaches provide a realistic description of the underlying radiation processes, their high computational cost makes them impractical for conventional parameter estimation methods such as Markov Chain Monte Carlo (MCMC). In this work, we introduce a surrogate model based on artificial neural networks (ANNs) that reproduces the output of a detailed one-zone kinetic SSC code with high accuracy. Combined with Posterior Estimation, this framework enables efficient simulation-based inference from broadband GRB afterglow observations while reducing the computational cost by several orders of magnitude. As a proof of concept, we apply our method to the TeV GRB 190114C and derive robust constraints on the microphysical parameters and the circumburst environment.