| Abstract: We develop a method for using a neural network to distinguish which stars fell onto the Milky Way versus those that we born with the galaxy. The network only uses kinematic input, which greatly extends the number of stars in the Gaia dataset which can be classified. The network is first trained on a cosmological zoom-in hydrodynamic simulated galaxy from the FIRE collaboration. A method known as transfer learning allows the network to train on a cross-matched Gaia/Rave dataset, which improvs the sensitivity to properties of the real Milky Way. Using the network’s classification of Gaia stars, we present evidence for a vast new stellar stream in the vicinity of the Sun. The kinematics of the stream stars are distinct from those of both the thin and thick disk. In particular, its rotational speed lags the disk by ~ 80 kms and its stars follow more eccentric orbits. A small number of Nyx stars have chemical abundances or inferred ages; from these, we deduce that the stream stars have a peak metallicity of [Fe/H] ~ -0.5 and ages 10-13 Gyr. Taken together with the kinematic observations, these results strongly favor the interpretation that Nyx is the remnant of a disrupted dwarf galaxy. |