EXO-200 releases public dataset for AI/ML applications

The EXO-200 collaboration has publicly released a subset of its calibration data, described in a new paper, EXO-200 Public Data Release for AI/ML Applications .

The dataset contains approximately one million events recorded during ²²⁸Th calibration campaigns near the end of EXO-200 Phase-II operations in early 2018. For every event it provides the raw waveforms from the U-wire, V-wire, and avalanche photodiode readout channels, together with reconstructed quantities such as charge energy, light energy, and charge-cluster information.

The data are distributed in HDF5 format and are intended to support data preservation, teaching, and the development of modern machine-learning techniques for rare-event physics.

• Paper: arXiv:2609.34304
• More on this line of research: EXO-200 & nEXO