IceCube citizen science challenge draws 4,000+ classifications to help train AI

CatAstro_Piyush · x · 2026-11-22

The IceCube Neutrino Observatory's "Name that Neutrino" citizen science challenge on Zooniverse received over 4,000 classifications from 65 volunteers worldwide, with top contributors announced. Participants classify neutrino signals as track or cascade shapes, helping eliminate background noise — their labels are compared against machine learning predictions to train and improve AI algorithms. The observatory records a terabyte of data daily, with a long-term goal of humans and AI working together to advance neutrino astronomy.

Original post →

More from Research

Research channel →