Research

Paper

AI LLM March 06, 2026

Transformer-Based Pulse Shape Discrimination in HPGe Detectors with Masked Autoencoder Pre-training

Authors

Marta Babicz, Saúl Alonso-Monsalve, Alain Fauquex, Laura Baudis

Abstract

Pulse-shape discrimination (PSD) in high-purity germanium (HPGe) detectors is central to rare-event searches such as neutrinoless double-beta decay (0vBB), yet conventional approaches compress each waveform into a small set of summary parameters, potentially discarding information in the full time series that is relevant for classification. We benchmark transformer-based models that operate directly on digitised waveforms using the Majorana Demonstrator AI/ML data release. Models are trained to reproduce the collaboration-provided accept/reject labels for four standard PSD cuts and to regress calibrated energy. We compare supervised training from scratch, masked autoencoder (MAE) self-supervised pre-training followed by fine-tuning, and a feature-based gradient-boosted decision tree (GBDT) baseline. Transformers outperform GBDT across all PSD targets, with the largest gains on the most challenging labels and on the combined PSD-pass definition. MAE pre-training improves sample efficiency, reducing labelled-data requirements by factors of 2-4 in low-label regimes. For energy regression, both transformer variants show a small common underestimation on the test split, while fine-tuning modestly narrows the residual distribution. These results motivate follow-up studies of robustness across detectors and operating conditions and of performance near QBB.

Metadata

arXiv ID: 2603.06192
Provider: ARXIV
Primary Category: hep-ex
Published: 2026-03-06
Fetched: 2026-03-09 06:05

Related papers

Raw Data (Debug)
{
  "raw_xml": "<entry>\n    <id>http://arxiv.org/abs/2603.06192v1</id>\n    <title>Transformer-Based Pulse Shape Discrimination in HPGe Detectors with Masked Autoencoder Pre-training</title>\n    <updated>2026-03-06T11:59:38Z</updated>\n    <link href='https://arxiv.org/abs/2603.06192v1' rel='alternate' type='text/html'/>\n    <link href='https://arxiv.org/pdf/2603.06192v1' rel='related' title='pdf' type='application/pdf'/>\n    <summary>Pulse-shape discrimination (PSD) in high-purity germanium (HPGe) detectors is central to rare-event searches such as neutrinoless double-beta decay (0vBB), yet conventional approaches compress each waveform into a small set of summary parameters, potentially discarding information in the full time series that is relevant for classification. We benchmark transformer-based models that operate directly on digitised waveforms using the Majorana Demonstrator AI/ML data release. Models are trained to reproduce the collaboration-provided accept/reject labels for four standard PSD cuts and to regress calibrated energy. We compare supervised training from scratch, masked autoencoder (MAE) self-supervised pre-training followed by fine-tuning, and a feature-based gradient-boosted decision tree (GBDT) baseline. Transformers outperform GBDT across all PSD targets, with the largest gains on the most challenging labels and on the combined PSD-pass definition. MAE pre-training improves sample efficiency, reducing labelled-data requirements by factors of 2-4 in low-label regimes. For energy regression, both transformer variants show a small common underestimation on the test split, while fine-tuning modestly narrows the residual distribution. These results motivate follow-up studies of robustness across detectors and operating conditions and of performance near QBB.</summary>\n    <category scheme='http://arxiv.org/schemas/atom' term='hep-ex'/>\n    <published>2026-03-06T11:59:38Z</published>\n    <arxiv:comment>21 pages, 9 figures</arxiv:comment>\n    <arxiv:primary_category term='hep-ex'/>\n    <author>\n      <name>Marta Babicz</name>\n    </author>\n    <author>\n      <name>Saúl Alonso-Monsalve</name>\n    </author>\n    <author>\n      <name>Alain Fauquex</name>\n    </author>\n    <author>\n      <name>Laura Baudis</name>\n    </author>\n  </entry>"
}