# You only thermoelastically deform once: Point Absorber Detection in LIGO Test Masses with YOLO

Source: [You only thermoelastically deform once: Point Absorber Detection in LIGO Test Masses with YOLO](<https://www.sota2.com/research/paper/68fa489d1e209131ce930e98>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-25T06:55:34.622Z

Authors: Simon R. Goode, Mitchell Schiworski, Daniel Brown, Eric Thrane, Paul D. Lasky

Published: 2024-11-25T05:32:27.000Z

## Abstract

Current and future gravitational-wave observatories rely on large-scale, precision interferometers to detect the gravitational-wave signals. However, microscopic imperfections on the test masses, known as point absorbers, cause problematic heating of the optic via absorption of the high-power laser beam, which results in diminished sensitivity, lock loss, or even permanent damage. Consistent monitoring of the test masses is crucial for detecting, characterizing, and ultimately removing point absorbers. We present a machine-learning algorithm for detecting point absorbers based on the object-detection algorithm You Only Look Once (YOLO). The algorithm can perform this task in situ while the detector is in operation. We validate our algorithm by comparing it with past reports of point absorbers identified by humans at LIGO. The algorithm confidently identifies the same point absorbers as humans with minimal false positives. It also identifies some point absorbers previously not identified by humans, which we confirm with human follow-up. We highlight the potential of machine learning in commissioning efforts.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2411.16104>)

[PDF](<https://arxiv.org/pdf/2411.16104.pdf>)
