Agents, chatbots and other tools based on artificial intelligence (AI) are increasingly used in everyday life by many. So-called large language model (LLM)-based agents, such as ChatGPT and Llama, have become impressively fluent in the responses they form, but quite often provide convincing yet incorrect information. Researchers at the University of Tokyo draw parallels between this issue and a human language disorder known as aphasia, where sufferers may speak fluently but make meaningless or hard-to-understand statements. This similarity could point toward better forms of diagnosis for aphasia, and even provide insight to AI engineers seeking to improve LLM-based agents.

Researchers from The University of Osaka have discovered the mechanism for supercurrent rectification, in which current flows primarily in one direction in a superconductor. By using a specific iron-based superconductor, they were able to observe this phenomenon over a broad range of magnetic and temperature fields. This understanding paves the way for the design and construction of superconducting diodes and other ultra-low energy electronics.

A research team from the Visual Perception and Cognition Laboratory and the Cognitive Neurotechnology Unit, Department of Computer Science and Engineering, Toyohashi University of Technology, Japan, investigated whether human pedestrians exhibit a consistent tendency to move either left or right to avoid autonomous mobile robots during head-on encounters. The results showed no consistent directional preference among the participants; instead, individual differences were observed in the chosen direction of avoidance. However, the direction in which the pedestrians moved, either left or right, could be reliably predicted by analyzing the orientation of their waist during walking. To respond accordingly, the robots were equipped with a function that enabled them to detect human waist orientation and adjust their movement in real time. This study also examined whether the timing of the robots' avoidance behavior influenced pedestrians’ comfort during passing. Overall, the findings indicated that early avoidance behavior by the robots, based on the predicted direction of human movement, led to an improvement in the pedestrians’ perceived comfort during passing. This study was published online in the journal PLoS One on May 14, 2025.

An Osaka Metropolitan University researcher investigated the home range and visiting areas of residents who live in Senboku NT.

Researchers at Toyohashi University of Technology in Japan, in collaboration with the Institute of Translational Medicine and Biomedical Engineering (IMTIB) in Argentina and the Indian Institute of Technology Madras, have advanced the "PDMS SlipChip," a versatile microfluidic device. By using a low-viscosity silicone oil and fine-tuning the fabrication process, they've made the SlipChip more reliable for cell-based experiments and simpler for creating concentration gradients. This breakthrough tackles previous issues like channel clogging and potential harm to cells, opening new avenues for biomedical research, including drug development and sophisticated cell studies.

Researchers from the RIKEN Center for Brain Science in Japan have discovered how inferred emotions are learned. The study shows that the frontal part of the brain coordinates with the amygdala—a brain region important for simple forms of emotional learning—to make this higher-order emotional ability possible. Published in the scientific journal Nature on May 14, this breakthrough study is the first to show how the brain codes human-like internal models of emotion.

Using a new fluorescent mouse model with advanced imaging techniques, researchers have successfully visualized how musculoskeletal components are integrated into the functional locomotor system during embryonic development.

Postpartum maternal mental health and mother-to-infant bonding are well-established as critical factors in a child’s psychosocial development. However, few studies have explored the combined impact of postpartum maternal depression and early bonding experiences on emotional and behavioral difficulties during middle childhood. A new study reveals significant associations between postpartum depression, mother-to-infant bonding, and child difficulties. Notably, secure early bonding was found to partially buffer the long-term effects of postpartum depression on child outcomes.

Semi-structured interviews conducted with bereaved family caregivers and healthcare professionals involved in end-of-life care show that family caregivers’ desires and healthcare professionals’ support are misaligned, leading to unintended, percolated work. Scientists investigated family caregivers’ experiences during their care journeys and the factors affecting cooperative activities between caregivers and medical professionals to understand the importance of collaboration in end-of-life care. Their findings propose a potential research trajectory towards medical professionals and technological support.

Researchers in Japan have developed a digital laboratory (dLab) system that fully automates the material synthesis and structural, physical property evaluation of thin-film samples. With dLab, the team can autonomously synthesize thin-film samples and measure their material properties. The team’s dLab system demonstrates advanced automatic and autonomous material synthesis for data- and robot-driven materials science.