This paper reports calculations that show how the angular polarization of thermal lepton pairs (dileptons) produced in high‑energy heavy‑ion collisions can carry direct information about the quark‑gluon plasma (QGP). The authors build a detailed, Lorentz‑covariant framework for dilepton emission. It combines virtual‑photon spectral functions computed at next‑to‑leading‑order (NLO) in the strong force and event‑by‑event iEBE‑MUSIC hydrodynamic simulations of the evolving fireball. They apply the framework to lead‑lead (Pb+Pb) collisions under conditions like those at the Large Hadron Collider (LHC).

At leading order (LO), dileptons come from quark–antiquark annihilation into a virtual photon that decays to a lepton pair. The authors include the NLO corrections that involve gluons. Those contributions cover Compton‑type scattering, modified annihilation with extra gluons, one‑loop corrections, and the Landau‑Pomeranchuk‑Migdal (LPM) effect, which changes radiation when many scatterings overlap. The paper concentrates on the intermediate‑mass region (IMR) of dilepton invariant mass, where partonic (quark and gluon) processes dominate and hadronic reactions are less important. The calculations produce angular distributions and a set of polarization coefficients (often denoted λθ, λφ, etc.) in several commonly used reference frames.

Why this matters: electromagnetic radiation such as photons and dileptons interacts only weakly with the hot medium, so it carries information from all stages of the collision. Dileptons have an extra handle compared with real photons: their invariant mass. That lets experiments probe different sources separately. The study shows that the polarization is driven by an in‑medium spectral function difference (called ρΔ), which vanishes in vacuum. In other words, nonzero polarization in the thermal signal points to medium effects inside the QGP. The authors also study how the polarization depends on the chosen measurement frame, on the assumed abundance of gluons before full thermal equilibrium (pre‑equilibrium), and they derive a one‑to‑one mapping between the polarization signals for dielectrons and dimuons. Such relations can help experiments compare the two decay channels.