Searcharxiv⌕ Search

arXiv subjects

Steve Croft

Publications and source records attributed to Steve Croft.

At least 19 recordsLinked to original sources
arXiv2026

Breakthrough Listen's Automated Commensal Technosignature Survey with MeerKAT

The search for extraterrestrial intelligence (SETI) is an ongoing effort to detect technosignatures, evidence of technologically capable life beyond Earth. Conducting a comprehensive SETI programme requires a large amount of telescope time, which must be balanced with the science goals of a given observatory. Fortunately, many modern radio telescopes offer commensal access to the data they produce, allowing multiple scientific programmes to operate in parallel. The MeerKAT radio telescope in South Africa provides commensal access to a range of components, from each antenna's digitiser to the main channeliser (F-engine), via multicast Ethernet groups. Here, we describe the Breakthrough Listen user-supplied equipment (BLUSE) system at MeerKAT, which leverages multicast Ethernet to conduct an autonomous commensal technosignature survey, processing the full available bandwidth from all antennas. Its primary mode of operation is to upchannelise the incoming F-engine data to ~1Hz resolution, synthesize coherent beams on objects of interest, and search the resultant data for technosignatures. Since 2022, BLUSE has autonomously processed data from coherent beams synthesized on more than 1.2 million individual pointings, including repeat visits. BLUSE demonstrates how commensal technosignature surveys on radio telescope arrays offer a rapid and cost-effective way to increase the rate at which technosignature surveys can be conducted. This article describes the architecture of BLUSE, provides experimental evidence validating its features and performance, and quantifies its observing progress over the past few years. We also discuss the technical evolution of BLUSE, examine challenges faced and addressed, and consider avenues for future research and development.

astro-ph.IM↗
arXiv2026

Searching for Extraterrestrial Intelligence with the SKA

The search for technosignatures (also known as the Search for Extraterrestrial Intelligence or SETI) depends critically on our ability to distinguish artificial signals from the rich complexity of natural astrophysical phenomena and radio frequency interference from anthropogenic emissions. As the search for technosignatures increasingly aligns with mainstream astrophysics, complementing the search for biosignatures, it demands not only sophisticated statistical and computational approaches, but also deep domain knowledge across the electromagnetic spectrum. The SKA will play a pivotal role in the next-generation of technosignature searches, providing an unprecedented combination of sensitivity, field of view, and spatial resolution over its wavelength range. Integrating wide-field, high-resolution observations with machine learning and multi-wavelength diagnostics will represent key steps forward. The SKA's singular capabilities will render it an indispensable instrument for the rapid identification and follow-up characterisation of promising technosignature candidates. In this chapter, we discuss the request for high temporal and spectral resolution data products with a main focus on frequency-domain SETI. Nevertheless, multiple SKA observing modes have the potential to substantially advance technosignature research.

astro-ph.IM↗
arXiv2026

Anomaly Hunter for Alerts (AHA): Anomaly Detection in the ZTF Transient Alert Stream

Modern time-domain surveys produce alert streams at a scale that makes exhaustive manual inspection infeasible, requiring automated methods to identify unusual transients for follow-up. In this work, we present an unsupervised anomaly detection pipeline applied to the ZTF alert stream using the Lasair broker. We define normal objects as SN Ia, SN II, and SN Ib/c. Anomalous objects include (i) more exotic transients (AGN, TDEs, SLSNe, CVs, and nuclear transients) and (ii) supernova-labeled objects, either spectroscopically or by Lasair, with anomalous properties, such as incorrect or absent host associations, or non-supernova-like light curves. Our pipeline consists of three independently trained simple autoencoders operating on distinct alert stream data products: object features, triplet image cutouts, and light curves. Each model is trained on predominantly normal transients, and performance is assessed using the recall of exotic objects and the purity of all anomalous objects across both a spectroscopically classified held-out test set and the live alert stream. In the test set, performance is evaluated at a fixed rank corresponding to the top ten scoring candidates, while in the alert stream it is evaluated using an anomaly threshold defined from test set behavior. Across both settings, the algorithms consistently recover exotic transients and anomalous supernovae among their top-ranked candidates. Over 25 days of live alert stream application, we identify 87 unusual supernova candidates for follow-up. The overlap between anomalies flagged by different autoencoders in the test set is non-existent, and in the alert stream is small, with maximum overlap between any two algorithms being 11 objects. The framework is data-efficient, requiring only a few thousand training examples, making it well suited for early and ongoing application to the Rubin Observatory alert stream.

astro-ph.SR↗
arXiv2026

On the Deepest Search for Galactic Center Pulsars and an Examination of an Intriguing Millisecond Pulsar Candidate

We report results of one of the most sensitive pulsar surveys to date targeting the innermost region of the Galactic Center (GC) using the Robert C. Byrd Green Bank Telescope (GBT) at X-band (8--12GHz) using data from the Breakthrough Listen initiative. In total, we collected 9.5 hr of data covering the wider $\sim 8'$ diameter of the GC bulge, and 11 hr on the inner $1.4'$ region between 2021 May and 2023 December. We conducted a comprehensive Fourier-domain periodicity search targeting both canonical pulsars (CPs) and millisecond pulsars (MSPs), using constant and linearly changing acceleration searches to improve sensitivity to compact binaries. Assuming weak scattering, our searches reached luminosity limits of $L_{\rm min} \approx 0.14~{\rm mJy~kpc^{2}}$ for CPs and $L_{\rm min} \approx 0.26~{\rm mJy~kpc^{2}}$ for MSPs -- sensitive enough to detect the most luminous pulsars expected in the GC. Among 5,282 signal candidates, we identify an interesting 8.19 ms MSP candidate (DM of 2775 pc cm$^{-3}$), persistent in time and frequency across a 1-hr scan at a flux density of $S_{\rm min} \approx 0.007~{\rm mJy}$. We introduce a novel randomization test for evaluating candidate significance against noise fluctuations, including signal persistence via Kolmogorov-Smirnov tests and flux-vs-DM behavior. We are unable to make a definitive claim about the candidate due to a mixed degree of confidence from these tests and, more broadly, its non-detection in subsequent observations. This deepens the ongoing missing pulsar problem in the GC, reinforcing the idea that strong scattering and/or extreme orbital dynamics may obscure pulsar signals in this region.

astro-ph.HE↗
arXiv2025

A Search for Radio Technosignatures from Interstellar Object 3I/ATLAS with the Allen Telescope Array

In 2025 July, the third-ever interstellar object, 3I/ATLAS, was discovered on its ingress into the Solar System. Similar to the NASA Voyager missions sent in 1977, science probes by extraterrestrial life ("artifact technosignatures") could be sent to explore other stellar systems like our own. In this campaign, we used the SETI Institute's Allen Telescope Array to observe 3I/ATLAS from 1-9 GHz. We detected nearly 74 million narrowband hits in 7.25\,hr of data using the newly-developed search pipeline bliss. We then blanked hits by frequency and drift rate to mitigate radio frequency interference in our dataset, narrowing the dataset down to ~2 million hits. These hits were further filtered by the localization code NBeamAnalysis, and the remaining 211 hits were visually inspected in the time-frequency domain. We did not find any signals worthy of additional follow-up. Accounting for the Doppler drift correction and given the non-detection, we are able to set an effective isotropic radiated power upper limit of 10-110 W on radio technosignatures from 3I/ATLAS across the frequency and drift rate ranges covered by our survey.

astro-ph.EP↗
arXiv2025

What We Don't C: Manifold Disentanglement for Structured Discovery

Accessing information in learned representations is critical for annotation, discovery, and data filtering in disciplines where high-dimensional datasets are common. We introduce What We Don't C, a novel approach based on latent flow matching that disentangles latent subspaces by explicitly removing information included in conditional guidance, resulting in meaningful residual representations. This allows factors of variation which have not already been captured in conditioning to become more readily available. We show how guidance in the flow path necessarily represses the information from the guiding, conditioning variables. Our results highlight this approach as a simple yet powerful mechanism for analyzing, controlling, and repurposing latent representations, providing a pathway toward using generative models to explore what we don't capture, consider, or catalog.

cs.AI↗
arXiv2025

Technosignature Searches of Interstellar Objects

With the discovery of the third confirmed interstellar object (ISO), 3I/ATLAS, we have entered a new phase in the exploration of these long-predicted objects. Though confirmed discovery of ISOs is quite recent, their utility as targets in the search for technosignatures (historically known as the Search for Extraterrestrial Intelligence -- SETI) has been discussed for many decades. With the upcoming NSF-DOE Vera C. Rubin Observatory's Legacy Survey of Space and Time (LSST), the discovery and tracking of such objects is expected to become routine, and thus so must our examination of these objects for possible technosignatures. Here we review the literature surrounding ISOs as targets for technosignatures, which provides a well-developed motivation for such exploration. We outline four broad classes of technosignatures that are well suited for ISO follow-up, including the type of data needed and the best timing for study. Given the limitations in the current understanding of ISOs, we show that care must be taken in identifying technosignatures based primarily on comparison to objects in the Solar System. We therefore provide a roadmap for careful and consistent study of the population of ISOs in the hope of identifying technosignatures.

astro-ph.IM↗
arXiv2025

Technosignature Searches with Real-time Alert Brokers

We present an exploration of technosignature research that is possible using real-time alert brokers from surveys such as the Zwicky Transient Facility (ZTF) and the upcoming Legacy Survey of Space and Time (LSST). Nine alert brokers currently stream up to 1 million alerts each night from ZTF, and LSST is projected to increase this volume by an order of magnitude. While these brokers are primarily designed to facilitate real-time follow-up of explosive transients such as supernovae, they offer a unique platform to discover rare forms of variability from nearby stars in real time, which is crucial for follow-up and characterization. We evaluate the capability for both spatial and temporal searches for extraterrestrial intelligence (SETI) methods using the currently available brokers, and present examples of technosignature searches using ZTF alert and archival data. We have deployed optical SETI techniques, such as planetary transit zone geometries and the SETI Ellipsoid. We have also developed a search for novel high-amplitude stellar dippers, and present a workflow that integrates features available directly through the brokers, as well as post-processing steps that build on the existing capabilities. Though the SETI methods that alert brokers can execute are currently limited, we provide suggestions that may enhance future technosignature and anomaly searches in the era of the Vera C. Rubin Observatory.

astro-ph.IM↗
arXiv2025

Using anomaly detection to search for technosignatures in Breakthrough Listen observations

We implement a machine learning algorithm to search for extra-terrestrial technosignatures in radio observations of several hundred nearby stars, obtained with the Parkes and Green Bank Telescopes by the Breakthrough Listen collaboration. Advances in detection technology have led to an exponential growth in data, necessitating innovative and efficient analysis methods. This problem is exacerbated by the large variety of possible forms an extraterrestrial signal might take, and the size of the multidimensional parameter space that must be searched. It is then made markedly worse by the fact that our best guess at the properties of such a signal is that it might resemble the signals emitted by human technology and communications, the main (yet diverse) contaminant in radio observations. We address this challenge by using a combination of simulations and machine learning methods for anomaly detection. We rank candidates by how unusual they are in frequency, and how persistent they are in time, by measuring the similarity between consecutive spectrograms of the same star. We validate that our filters significantly improve the quality of the candidates that are selected for human vetting when compared to a random selection. Of the ~ 10^11 spectrograms that we analyzed, we visually inspected thousands of the most promising spectrograms, and thousands more for validation, about 20,000 in total, and report that no candidate survived basic scrutiny.

astro-ph.IM↗
arXiv2024

A Novel Technosignature Search in the Breakthrough Listen Green Bank Telescope Archive

The Breakthrough Listen program is, to date, the most extensive search for technological life beyond Earth. As part of this goal, over the past nine years it has surveyed thousands of nearby stars, close to 100 nearby galaxies, and a variety of exotic and solar system objects with telescopes around the world, including the Robert C. Byrd Green Bank Telescope (GBT) in West Virginia. The goal is to find evidence of technosignatures of other civilizations, such as narrowband Doppler drifting radio signals. Despite the GBT's location in a radio-quiet zone, the primary challenge of this search continues to be the high quantities of human-generated radio-frequency interference (RFI), and the ability to pick out genuinely promising candidates from it. Here we present a novel search method aimed at finding these `needle-in-a-haystack' type signals, applied to 6,630 observation cadences of 2,623 stars (each observed with one or more of the L, S, C, and X band receivers) from the GBT archive. We implement a low-complexity statistical process to vet out RFI and highlight signals that, upon visual inspection, appear more promising than those from previous analyses. Our work returns candidate signals found previously using both traditional and machine learning algorithms, as well as many promising ones not previously identified. This analysis represents the largest dataset searched for technosignatures to date, and highlights the efficacy that traditional (non-machine-learning) algorithms continue to have in these types of technosignature searches. We find that less than 1% of stars host transmitters brighter than 0.3 Arecibo radar equivalents broadcasting in our direction over the frequency band covered.

astro-ph.IM↗
arXiv2024

Anomaly Detection and Radio-frequency Interference Classification with Unsupervised Learning in Narrowband Radio Technosignature Searches

The search for radio technosignatures is an anomaly detection problem: Candidate signals represent needles of interest in the proverbial haystack of radio-frequency interference (RFI). Current search frameworks find an enormity of false-positive signals, especially in large surveys, requiring manual follow-up to a sometimes prohibitive degree. Unsupervised learning provides an algorithmic way to winnow the most anomalous signals from the chaff, as well as group together RFI signals that bear morphological similarities. We present GLOBULAR (Grouping Low-frequency Observations By Unsupervised Learning After Reduction) clustering, a signal processing method that uses HDBSCAN to reduce the false-positive rate and isolate outlier signals for further analysis. When combined with a standard narrowband signal detection and spatial filtering pipeline, such as turboSETI, GLOBULAR clustering offers significant improvements in the false-positive rate over the standard pipeline alone, suggesting dramatic potential for the amelioration of manual follow-up requirements for future large surveys. By removing RFI signals in regions of high spectral occupancy, GLOBULAR clustering may also enable the detection of signals missed by the standard pipeline. We benchmark our method against the Choza et al. turboSETI-only search of 97 nearby galaxies at the L band, demonstrating a false-positive hit reduction rate of 93.1% and a false-positive event reduction rate of 99.3%.

astro-ph.IM↗
arXiv2024

The First High Frequency Technosignature Search Survey with the Sardinia Radio Telescope

The quest for radio signals from technologically-advanced extraterrestrial intelligence has traditionally concentrated on the vicinity of 1.4 GHz. In this paper, we extend the search to unprecedented territories, detailing our extensive observations at 6 GHz and initiating the first-ever survey at 18 GHz with the Sardinia Radio Telescope (SRT). Our strategy entailed rigorous observation sessions, totaling 36 hours, directed towards the Galactic Center and 72 selected TESS targets-making this the most comprehensive high-frequency technosignature search to date. Our narrowband signal search found no definitive evidence of drifting signals that could suggest an extraterrestrial origin from the surveyed regions. Nevertheless, our efforts have enabled us to set new constraints on the presence of radio emissions from approximately $5\times 10^{5}$ stars, establishing an isotropic radiated power limit of $1.8\times 10^{19} W$. We also provide a comparative analysis of the 'hits' recorded across both frequencies to highlight the significance of pursuing technosignature searches at higher frequencies, where the spectral landscape is less congested and more conducive to detection.

astro-ph.HE↗
arXiv2024

Searching the SN 1987A SETI Ellipsoid with TESS

The SETI Ellipsoid is a strategy for technosignature candidate selection which assumes that extraterrestrial civilizations who have observed a galactic-scale event -- such as supernova 1987A -- may use it as a Schelling point to broadcast synchronized signals indicating their presence. Continuous wide-field surveys of the sky offer a powerful new opportunity to look for these signals, compensating for the uncertainty in their estimated time of arrival. We explore sources in the TESS continuous viewing zone, which corresponds to 5% of all TESS data, observed during the first three years of the mission. Using improved 3D locations for stars from Gaia Early Data Release 3, we identified 32 SN 1987A SETI Ellipsoid targets in the TESS continuous viewing zone with uncertainties better than 0.5 ly. We examined the TESS light curves of these stars during the Ellipsoid crossing event and found no anomalous signatures. We discuss ways to expand this methodology to other surveys, more targets, and different potential signal types.

astro-ph.IM↗
arXiv2024

The weird and the wonderful in our Solar System: Searching for serendipity in the Legacy Survey of Space and Time

We present a novel method for anomaly detection in Solar System object data, in preparation for the Legacy Survey of Space and Time. We train a deep autoencoder for anomaly detection and use the learned latent space to search for other interesting objects. We demonstrate the efficacy of the autoencoder approach by finding interesting examples, such as interstellar objects, and show that using the autoencoder, further examples of interesting classes can be found. We also investigate the limits of classic unsupervised approaches to anomaly detection through the generation of synthetic anomalies and evaluate the feasibility of using a supervised learning approach. Future work should consider expanding the feature space to increase the variety of anomalies that can be uncovered during the survey using an autoencoder.

astro-ph.EP↗
arXiv2023

The Breakthrough Listen Search for Intelligent Life: Detection and Characterization of Anomalous Transits in Kepler Lightcurves

Never before has the detection and characterization of exoplanets via transit photometry been as promising and feasible as it is now, due to the increasing breadth and sensitivity of time domain optical surveys. Past works have made use of phase-folded stellar lightcurves in order to study the properties of exoplanet transits, because this provides the highest signal that a transit is present at a given period and ephemeris. Characterizing transits on an individual, rather than phase-folded, basis is much more challenging due to the often low signal-to-noise ratio (SNR) of lightcurves, missing data, and low sampling rates. However, by phase-folding a lightcurve we implicitly assume that all transits have the same expected properties, and lose all information about the nature and variability of the transits. We miss the natural variability in transit shapes, or even the deliberate or inadvertent modification of transit signals by an extraterrestrial civilization (for example, via laser emission or orbiting megastructures). In this work, we develop an algorithm to search stellar lightcurves for individual anomalous (in timing or depth) transits, and we report the results of that search for 218 confirmed transiting exoplanet systems from Kepler.

astro-ph.EP↗
arXiv2023

The Breakthrough Listen Search for Intelligent Life: Technosignature Search of 97 Nearby Galaxies

The Breakthrough Listen search for intelligent life is, to date, the most extensive technosignature search of nearby celestial objects. We present a radio technosignature search of the centers of 97 nearby galaxies, observed by Breakthrough Listen at the Robert C. Byrd Green Bank Telescope. We performed a narrowband Doppler drift search using the turboSETI pipeline with a minimum signal-to-noise parameter threshold of 10, across a drift rate range of $\pm$ 4 Hz\ $s^{-1}$, with a spectral resolution of 3 Hz and a time resolution of $\sim$ 18.25 s. We removed radio frequency interference by using an on-source/off-source cadence pattern of six observations and discarding signals with Doppler drift rates of 0. We assess factors affecting the sensitivity of the Breakthrough Listen data reduction and search pipeline using signal injection and recovery techniques and apply new methods for the investigation of the RFI environment. We present results in four frequency bands covering 1 -- 11 GHz, and place constraints on the presence of transmitters with equivalent isotropic radiated power on the order of $10^{26}$ W, corresponding to the theoretical power consumption of Kardashev Type II civilizations.

astro-ph.IM↗
arXiv2023

Developing a Drift Rate Distribution for Technosignature Searches of Exoplanets

A stable-frequency transmitter with relative radial acceleration to a receiver will show a change in received frequency over time, known as a "drift rate''. For a transmission from an exoplanet, we must account for multiple components of drift rate: the exoplanet's orbit and rotation, the Earth's orbit and rotation, and other contributions. Understanding the drift rate distribution produced by exoplanets relative to Earth, can a) help us constrain the range of drift rates to check in a Search for Extraterrestrial Intelligence (SETI) project to detect radio technosignatures and b) help us decide validity of signals-of-interest, as we can compare drifting signals with expected drift rates from the target star. In this paper, we modeled the drift rate distribution for $\sim$5300 confirmed exoplanets, using parameters from the NASA Exoplanet Archive (NEA). We find that confirmed exoplanets have drift rates such that 99\% of them fall within the $\pm$53 nHz range. This implies a distribution-informed maximum drift rate $\sim$4 times lower than previous work. To mitigate the observational biases inherent in the NEA, we also simulated an exoplanet population built to reduce these biases. The results suggest that, for a Kepler-like target star without known exoplanets, $\pm$0.44 nHz would be sufficient to account for 99\% of signals. This reduction in recommended maximum drift rate is partially due to inclination effects and bias towards short orbital periods in the NEA. These narrowed drift rate maxima will increase the efficiency of searches and save significant computational effort in future radio technosignature searches.

astro-ph.EP↗