Toggle light / dark theme

In an era where data privacy concerns loom large, a new approach in artificial intelligence (AI) could reshape how sensitive information is processed.

Researchers Austin Ebel and Karthik Garimella, Ph.D. students, and Assistant Professor of Electrical and Computer Engineering Brandon Reagen have introduced Orion, a novel framework that brings fully (FHE) to deep learning—allowing AI models to practically and efficiently operate directly on encrypted data without needing to decrypt it first.

The implications of this advancement, published on the arXiv preprint server and scheduled to be presented at the 2025 ACM International Conference on Architectural Support for Programming Languages and Operating Systems, are profound.

Researchers at the UAB have developed a new chemical reaction to form solid polymeric networks using light (photocuring) which will allow the preparation of solid materials with controlled shapes measuring under a thousandth of a millimeter. The research is key for the development of new, performance-enhanced lithographic and 3D printing techniques.

At present, 3D printing is an increasingly widespread and accessible technology, typically involving the formation of solid polymeric materials in a specific region, either by extruding pre-formed polymers or by generating them in situ from their corresponding monomers, the molecules that make up polymers.

However, these techniques often suffer from several drawbacks, such as long printing times or low resolution, preventing the production of printed materials with micrometric dimensions.

For decades, researchers have explored how electrons behave in quantum materials. Under certain conditions, electrons interact strongly with each other instead of moving independently, leading to exotic quantum states. One such state, first proposed by Nobel laureate Eugene Wigner, is the Wigner crystal—a structured electron arrangement caused by their mutual repulsion. Although widely theorized, experimental proof has been rare.

Researchers at Yonsei University have now provided evidence of Wigner crystallization and the associated electronic rotons. In a study published in the journal Nature, Prof. Keun Su Kim and his team used (ARPES) to analyze black phosphorus doped with alkali metals. Their data revealed aperiodic energy variations, a hallmark of electronic rotons.

Crucially, as they decreased the dopant density within the material, the roton energy gap shrank to zero. This observation confirmed a transition from a fluid-like quantum state to a structured electron lattice, characteristic of Wigner crystallization.

Researchers from the German Primate Center—Leibniz Institute for Primate Research and the Max Planck Institute of Molecular Cell Biology and Genetics have discovered two specific genes that evolve exclusively in humans jointly influence the development of the cerebrum. They have thus provided evidence that these genes contribute together to the evolutionary enlargement of the brain.

The work has been published in Science Advances.

The results show that the two genes act in a finely tuned interplay: one ensures that the progenitor cells of the brain multiply more, while the other causes these cells to transform into a different type of progenitor cell—the cells that later form the nerve cells of the brain. In the course of evolution, this interplay has led to the being unique in its size and complexity.

Everything the brain does—from storing memories to interpreting sights to regulating emotions—requires energy, all produced by cellular organelles called mitochondria.

However, surprisingly little is known about the distribution and diversity of the brain’s tiny energy processors and how they influence brain health. For instance, how many mitochondria does the brain have? Are they uniformly distributed across the whole brain? Are all brain mitochondria the same? Do changes in the brain’s mitochondria affect mood, cognition, and the development of neurological and psychiatric conditions?

To begin answering these and other questions, Columbia University researchers have created MitoBrainMap, the first-ever atlas of the brain’s mitochondria.

Highly charged heavy ions form a very suitable experimental field for investigating quantum electrodynamics (QED), the best-tested theory in physics describing all electrical and magnetic interactions of light and matter. A crucial property of the electron within QED is the so-called g factor, which precisely characterizes how the particle behaves in a magnetic field.

Recently, the ALPHATRAP group led by Sven Sturm in the division of Klaus Blaum at the Max-Planck-Institut für Kernphysik (MPIK) in Heidelberg measured the g factor of hydrogen-like tin ions on a precision level of 0.5 parts per billion, which is like measuring the distance from Cologne to Frankfurt with precision down to the thickness of a human hair. This is a stringent test of QED for the simplest atomic system, just like conventional hydrogen but with a much higher electric field experienced by the electron due to the charge of 50 protons inside the tin nucleus.

In a new study published in Physical Review Letters, researchers have now tackled highly charged boron-like tin ions with only five remaining electrons. The goal is to study the inter-electronic effects in the boron-like configuration. So far, the only boron-like g factor has been measured with high precision for argon ions with a proton number Z of 18. However, the nucleus is not a point charge like the electron and its charge distribution leads to finite nuclear size corrections—another challenge for precision experiments.

Georgia Tech researchers recently proposed a method for generating quantum entanglement between photons. This method constitutes a breakthrough that has potentially transformative consequences for the future of photonics-based quantum computing.

“Our results point to the possibility of building quantum computers using light by taking advantage of this entanglement,” said Chandra Raman, a professor in the School of Physics. The research is published in the journal Physical Review Letters.

Quantum computers have the potential to outperform their conventional counterparts, becoming the fastest programmable machines in existence. Entanglement is the key resource for building these quantum computers.

On March 24, at the annual Rencontres de Moriond conference taking place in La Thuile, Italy, the LHCb collaboration at CERN reported a new milestone in our understanding of the subtle yet profound differences between matter and antimatter.

In its analysis of large quantities of data produced by the Large Hadron Collider (LHC), the international team found overwhelming evidence that particles known as baryons, such as the protons and neutrons that make up , are subject to a mirror-like asymmetry in nature’s fundamental laws that causes matter and antimatter to behave differently.

The discovery provides new ways to address why the that make up matter fall into the neat patterns described by the Standard Model of particle physics, and to explore why matter apparently prevailed over antimatter after the Big Bang. The paper is available on the arXiv preprint server.

Similar to humans going on journeys of self-discovery, quantum computers are also capable of deepening their understanding of their own foundations.

Researchers from Tohoku University and St. Paul’s School, London, have developed a that allows quantum computers to analyze and protect quantum entanglement—a fundamental underpinning of quantum computing. These findings will advance our understanding of quantum entanglement and quantum technologies.

The study was published in Physical Review Letters on March 4, 2025.

At hypersonic speeds, complexities occur when the gases interact with the surface of the vehicle, such as boundary layers and shock waves. Researchers in the Department of Aerospace Engineering at The Grainger College of Engineering, University of Illinois Urbana-Champaign, were able to observe new disturbances in simulations conducted for the first time in 3D.

The study, “Loss of axial symmetry in hypersonic flows over conical shapes,” is published in Physical Review Fluids.

Fully 3D simulations require a great deal of processing power, making the work expensive to compute. Two things made it possible for Deborah Levin and her Ph.D. student Irmak Taylan Karpuzcu to conduct the research: Time on Frontera, the leadership-class computer system at the Texas Advanced Computing Center and software developed in previous years by several of Levin’s former graduate students.