Neural posterior estimation of the neutrino direction in IceCube using transformer-encoded normalizing flows on the sphere

IceCube Collaboration; R Abbasi; M Ackermann; J Adams; J. A Aguilar; M Ahlers; J M Alameddine; S Ali; N. M Amin; M Hostert

doi:10.48550/arxiv.2604.19846

Back

Neural posterior estimation of the neutrino direction in IceCube using transformer-encoded normalizing flows on the sphere

Preprint

Open access

Neural posterior estimation of the neutrino direction in IceCube using transformer-encoded normalizing flows on the sphere

IceCube Collaboration, R Abbasi, M Ackermann, J Adams, J. A Aguilar, M Ahlers, J M Alameddine, S Ali, N. M Amin and M Hostert

ArXiv.org

Cornell University

04/21/2026

DOI: 10.48550/arxiv.2604.19846

Files and links (1)

url

https://doi.org/10.48550/arxiv.2604.19846View

Preprint (Author's original) This preprint has not been evaluated by subject experts through peer review. Preprints may undergo extensive changes and/or become peer-reviewed journal articles. Open Access

Abstract

IceCube is a cubic-kilometer-scale neutrino detector located at the geographic South Pole. A precise directional reconstruction of IceCube neutrinos is vital for associations with astronomical objects. In this context, we discuss neural posterior estimation of the neutrino direction via a transformer encoder that maps to a normalizing flow on the 2-sphere. It achieves a new state-of-the-art angular resolution for the two main event morphologies in IceCube - tracks and showers - while being significantly faster than traditional B-spline-based likelihood reconstructions. All-sky scans can be performed within seconds rather than hours, and take constant computation time, regardless of whether the posterior extent is arc-minutes or spans the whole sky. We utilize a combination ofC² -smooth rational-quadratic splines, scale transformations and rotations to define a novel spherical normalizing-flow distribution whose parameters are predicted as a whole as the output of the transformer encoder. We test several structural choices diverting from the vanilla transformer architecture. In particular, we find dual residual streams, nonlinear QKV projection and a separate class token with its own cross-attention processing to boost test-time performance. The angular resolution for both showers and tracks improves substantially over the whole trained energy range from 100 GeV to 100 PeV. At 100 TeV deposited energy, for example, the median angular resolution improves by a factor of1.3for throughgoing tracks, by a factor of1.7for showers and by a factor of2.5for starting tracks compared to state-of-the art likelihood reconstructions based on B-splines. While previous machine-learning (ML) efforts have managed to obtain competitive shower resolutions, this is the first time an ML-based method outperforms likelihood-based muon reconstructions above 100 GeV.

Computer Science - Artificial Intelligence

Computer Science - Learning

Physics - High Energy Astrophysical Phenomena

Physics - High Energy Physics - Experiment

Physics - Instrumentation and Methods for Astrophysics

Details

Title: Subtitle: Neural posterior estimation of the neutrino direction in IceCube using transformer-encoded normalizing flows on the sphere
Creators: IceCube Collaboration
R Abbasi (Contributor) - Loyola University Chicago
M Ackermann (Contributor) - Deutsches Elektronen-Synchrotron DESY
J Adams (Contributor) - University of Canterbury
J. A Aguilar (Contributor) - Université Libre de Bruxelles
M Ahlers (Contributor) - University of Copenhagen
J M Alameddine (Contributor) - TU Dortmund University
S Ali (Contributor) - University of Kansas
N. M Amin (Contributor) - University of Delaware
M Hostert (Contributor)
Resource Type: Preprint
Publication Details: ArXiv.org
DOI: 10.48550/arxiv.2604.19846
ISSN: 2331-8422
Publisher: Cornell University; Ithaca, New York
Language: English
Date posted: 04/21/2026
Academic Unit: Physics and Astronomy
Record Identifier: 9985156523802771

Metrics

1 Record Views