# A deep-learning search for technosignatures of 820 nearby stars

**Authors:** Peter Xiangyuan Ma, Cherry Ng, Leandro Rizk, Steve Croft, Andrew P. V. Siemion, Bryan Brzycki, Daniel Czech, Jamie Drew, Vishal Gajjar, John Hoang, Howard Isaacson, Matt Lebofsky, David MacMahon, Imke de Pater, Danny C. Price, Sofia Z. Sheikh, S. Pete Worden  
**arXiv:** [2301.12670](https://arxiv.org/abs/2301.12670)  
**Date:** 2023-01-30  
**Summary page:** https://www.summarizepaper.com/en/arxiv-id/2301.12670v1/

## Key points

- Scientists search for extraterrestrial intelligence (SETI) by detecting "technosignatures"
- Narrowband Doppler drifting radio signals are one type of technosignature
- Radio frequency interference (RFI) poses a significant challenge to SETI in the radio domain
- A team of researchers led by Peter Xiangyuan Ma conducted a deep-learning based technosignature search and found eight promising ETI signals
- The study involved observing 820 unique targets with the Robert C. Byrd Green Bank Telescope, totaling over 480 hours of on-sky data
- The researchers used a novel beta-Convolutional Variational Autoencoder to identify technosignature candidates in a semi-unsupervised manner while keeping the false positive rate manageable
- This new approach presents itself as a leading solution in accelerating SETI and other transient research into the age of data-driven astronomy

## Summary

In the search for extraterrestrial intelligence (SETI), scientists aim to identify the prevalence of technological life beyond Earth by detecting their "technosignatures". One such technosignature is narrowband Doppler drifting radio signals. However, conducting SETI in the radio domain poses a significant challenge due to human radio frequency interference (RFI). To address this issue, a team of researchers led by Peter Xiangyuan Ma conducted a deep-learning based technosignature search which has returned eight promising ETI signals of interest for re-observation as part of the Breakthrough Listen initiative. The study involved observing 820 unique targets with the Robert C. Byrd Green Bank Telescope, totaling over 480 hours of on-sky data. The researchers implemented a novel beta-Convolutional Variational Autoencoder to identify technosignature candidates in a semi-unsupervised manner while keeping the false positive rate manageable. This new approach presents itself as a leading solution in accelerating SETI and other transient research into the age of data-driven astronomy. The study was published in Nature Astronomy and authored by Cherry Ng, Leandro Rizk, Steve Croft, Andrew P.V. Siemion, Bryan Brzycki, Daniel Czech, Jamie Drew, Vishal Gajjar, John Hoang, Howard Isaacson, Matt Lebofsky , David MacMahon Imke de Pater Danny C. Price Sofia Z. Sheikh and S.Pete Worden alongside Peter Xiangyuan Ma.