What’s known

The algorithm was developed by researchers David O’Ryan and Pablo Gómez. It was trained on the Hubble Legacy Archive, a database containing tens of thousands of observational datasets collected over 35 years of the telescope’s operation. ESA scientists note that the sheer volume of data makes detailed manual analysis practically impossible.

After an initial automated screening, the results were reviewed by astronomers. In total, the system identified around 1,400 anomalous objects, more than 800 of which had never been documented before. Most of the discoveries include interacting or merging galaxies, as well as gravitational lenses, where the light from distant objects is bent by massive foreground galaxies.

Researchers also detected protoplanetary disks, “jellyfish” galaxies, and objects featuring unusual star cluster formations. Several dozen findings could not be classified into any known category.

According to the project’s authors, AnomalyMatch could become a powerful tool for analyzing massive astronomical archives in the future, significantly accelerating discoveries and enabling scientists to uncover rare and unexpected cosmic phenomena.