Scientists from Vilnius University (VU) have made significant progress in understanding rare, hot subdwarf stars by using artificial intelligence (AI) and data from the Gaia space observatory. Hot subdwarfs are small, extremely hot stars that do not fit into the typical categories of stars. According to the latest Gaia data release, there are about 60,000 potential hot subdwarf stars among the over 500 million stars cataloged by Gaia, a European Space Agency mission that maps the Milky Way. Hot subdwarfs are unusual because they are much hotter than most stars of their size and do not align with the standard star classification diagram, which plots temperature against brightness. Most stars, including our sun, lie along the main sequence, a region that represents stable, hydrogen-burning stars. In contrast, hot subdwarfs are found below this main sequence and are often described as exposed stellar cores. These stars are typically formed when a companion star strips away the outer layers of a more massive star, causing it to evolve into a white dwarf in a nonstandard way. To identify these rare stars, the research team trained a convolutional neural network, a type of AI that excels at recognizing patterns in data. The AI was trained on 2,500 stars with known classifications and then used to analyze 17,500 other stars in the Gaia dataset. The AI detected patterns in the Gaia XP spectra data that match existing physical theories, confirming current scientific understanding through machine learning. Importantly, the AI identified these patterns without prior knowledge of physics, demonstrating the power of machine learning in astrophysical research. The project brought together researchers from different disciplines and countries. Dr. Aidas Medžiūnas from Vilnius University’s Faculty of Mathematics and Informatics played a key role in optimizing the use of the Gaia XP spectra database. Professor Ana Ulla from the University of Vigo, who has studied hot subdwarf stars for over 30 years, also contributed to the research and emphasized the importance of international collaboration and diverse expertise. The team’s work has already led to two scientific papers, two grants, and the first observations of hot subdwarfs from Lithuania at the VU Molėtai Astronomical Observatory. The findings are published in the journal Astronomy & Astrophysics, and the team plans to continue their research with the upcoming release of Gaia DR4 in December and an expanding international team.