A once-theoretical structure is now shaping the future of computing. Researchers have uncovered key properties of magnetic skyrmions, ultra-stable, 2 nm vortex-like structures that could enable next-gen memory with extremely low power consumption. Even more surprising? They can form in materials once thought impossible.

Researchers are investigating how 3D printers could achieve unprecedented levels of precision when creating batteries, opening the door for new innovations.

How can public spaces remain safe when large crowds move through them? Engineers and researchers who study these environments often rely on physical models borrowed from fluid dynamics — a branch of physics that describes the collective motion of fluids, whose behaviour emerges from the interactions of many particles. But a new study published in the Journal of Statistical Physics: Theory and Experiment (JSTAT) highlights a crucial issue: the way data are collected and measured within these models lacks standardisation and may overlook important features of human collective behaviour. Unlike particles, people are living agents with individual decisions and complex interactions, making their movement harder to capture with traditional approaches. In their study, the authors propose and experimentally test new methods to address these limitations, comparing them with more established techniques. Their results point the way toward clearer methodological guidelines and the development of more reliable tools for those involved in the design and management of public spaces.

Researchers in Japan have developed a new LED structure that generates circularly polarized light from a single chip. The advance could support smaller and more energy-efficient optical devices for AR/VR, 3D displays, quantum communication, and optical security. By combining a semipolar InGaN light-emitting structure with a stripe-shaped silicon nitride metasurface, the team created a compact light source that reduces energy-conversion loss and operates at room temperature. This advancement could help bring ultra-compact, durable light sources closer to practical use in AR/VR, 3D displays, quantum communication, and optical security.

Researchers analyzed communication across the animal kingdom, including firefly flashes, cricket chirps, frog croaks, birds’ mating displays and more. Across species, many communication signals repeat at two beats per second. Brains are most effective at processing signals that arrive about twice per second. Findings suggest communication signals may have evolved to match the rhythms the brain processes most easily.

The Subaru Telescope observed the interstellar comet 3I/ATLAS (C/2025 N1) on January 7, 2026 (UT), after it made its closest approach to the Sun. By observing colors in the coma around the comet, astronomers could estimate the ratio of carbon dioxide to water. This ratio is much lower than that inferred from earlier observations by space telescopes. These findings suggest that the chemistry of the coma is evolving over time and offers clues to the structure of comet 3I/ATLAS.

Insilico Medicine has expanded its Science MMAI Gym, a large-scale training and benchmarking platform for artificial intelligence, with the launch of three public leaderboard portals designed to evaluate AI performance across scientific research and drug discovery. Positioned as both a training environment and benchmarking system, MMAI Gym enables the development of domain-specific AI models while rigorously assessing their capabilities on real-world tasks.

The newly launched benchmark categories include ScienceAI Bench, which evaluates general scientific reasoning across disciplines such as biology, chemistry, and materials science; the Drug Discovery Benchmark (DDB), focused on end-to-end pharmaceutical R&D tasks; and Insilico Bench, a proprietary suite targeting complex and emerging scientific challenges. Together, these benchmarks draw from both curated industry datasets and proprietary, experimentally grounded data, enabling multi-dimensional evaluation across more than 200 tasks.

The platform reflects a broader shift toward standardized, scalable evaluation of scientific AI systems. Previous results from Insilico demonstrate that models trained within MMAI Gym can achieve up to tenfold performance improvements on key drug discovery benchmarks. In collaboration with Liquid AI, Insilico also developed a compact foundation model that achieved state-of-the-art performance across multiple drug discovery tasks, with findings presented at ICLR 2026.

By integrating training, benchmarking, and public evaluation, MMAI Gym aims to accelerate the adoption of reliable, high-performance AI systems across pharmaceutical research and beyond.

New hydrodynamical simulations explain how tidally destroyed stars reveal hidden supermassive black holes and why no two of these cosmic collisions look the same.