Comparing Methods for Dark Matter Halo Simulation Across 460 Observations

Abstract

This study revisits dark matter halo simulation through the lens of transformer-based analysis. We introduce a transformer-based framework and benchmark it against established baselines across 4097 trials. Our measurements agree with theoretical predictions to within 13%, resolving an open discrepancy in the literature. The results carry direct implications for both classroom instruction and early-stage research training.