Parts | Days | Selection | Search | Updates | Downloads | Help

GR: Fachverband Gravitation, Relativistische Astrophysik und Kosmologie

GR 7: Relativistic Astrophysics II

GR 7.2: Talk

Wednesday, March 18, 2026, 11:30–11:45, KH 01.016

Construction of a 2d neutrino sky map — •Lukas Liland for the IceCube collaboration — TU Dortmund

The IceCube collaboration has during the last 15 years made important discoveries concerning high-energy astrophysical neutrinos. IceCube first presented evidence of the flux of these neutrinos in 2013, and in 2023 the collaboration presented evidence of an excess neutrino flux from the Milky Way Galaxy. This talk presents a project with the aim of constructing a 2D sky map of the astrophysical neutrino flux. A relatively new sample of IceCube neutrino events are processed with a transformer neural network, a state-of-the-art machine learning algorithm used by the famous large language models, to yield improved reconstructions of the energies and the directions of the neutrinos. A neutrino sky map is then produced using the technique of unfolding, which recovers the real neutrino flux from the calculated estimations with the help of MC simulations.

Keywords: neutrinos; unfolding; neural networks; machine learning