Reliable Parameter Inference for the Epoch of Reionization using Balanced Neural Ratio Estimation

We present an application of the Balanced Neural Ratio Estimation algorithm to improve the statistical validity of parameter estimates used to characterize the Epoch of Reionization, where the common assumption of a multivariate Gaussian likelihood leads to overconfident posterior distributions. Our results demonstrate the potential of Simulation-Based Inference (SBI) methods to provide statistically robust constraints, and highlight how easily they can be adopted within existing astrophysical modeling frameworks.