Quantum computers—systems that process information and perform computations by leveraging the principles of quantum mechanics—could solve some tasks faster and more effectively than classical computers. While some studies have demonstrated the advantages of these computers for specific tasks, ensuring their reliable operation in real-world settings has proved challenging.

A study in the Journal of Cosmology and Astroparticle Physics explores how a machine-learning strategy known as transfer learning could dramatically reduce the computational cost of searching for new physics beyond the standard cosmological model—while also revealing an unexpected risk: Sometimes AI systems can become too reliant on what they already know.

How are heavy elements formed in the universe? Extremely neutron-rich atomic nuclei and their beta-decay rates play an important role in this process. Until now, it has been very difficult to determine these rates experimentally. Researchers at TU Darmstadt have developed theoretical predictions for such processes and successfully compared them with experimental data, where they exist. The results were published in Physical Review Letters.

Nature is filled with remarkable visual phenomena created by microscopic surface structures that interact with light in fascinating ways. The iridescent wings of butterflies, the shimmering feathers of birds and the glossy surfaces of flower petals are all examples of how living organisms control the reflection, absorption and scattering of light. These optical effects are not only visually striking but also serve important biological functions, including attracting pollinators, communication, camouflage and protection from environmental stress. Understanding these naturally occurring photonic structures has become an important area of research, as they provide inspiration for the development of advanced biomimetic materials and optical technologies.