Hypertriton mass measurement in PbPb collisions at 5.02 TeV. The hypertriton signal is extracted using Machine Learning models (Boosted Decision Trees) to improve the signal/background discrimination. The mass values are extracted from the invariant mass fit where a Kernel Density Estimation based on MC data is used to describe the signal p.d.f. The blue line represents the pol0 fit to the masses and the shaded area the uncertainties, statistical in dark blue and systematic in light blue.