Torres-Quijano et al. used machine learning to predict which stars are likely to host small planets based on their chemical composition. Their model identified sodium (Na) and vanadium (V) as key indicators, outperforming iron (Fe). The study validated its predictions and suggested that future exoplanet searches, including NASA missions, could use these findings to improve planet detection efficiency. This research advances our understanding of planetary formation and the star-planet connection.