A neural-network potential developed in Japan enables simulations of 100,000-atom silicon crystals, revealing that hydrogen-induced platelets contain two to three...

A dynamical system deep learning framework grounded in Takens' embedding theorem recovers large-scale climate predictability from sparse observations, skillfully hindcasting...

Researchers at Jilin University have developed a Kolmogorov-Arnold network-based inverse mapping model that identifies groundwater contamination sources within seconds, achieving...

Researchers have developed a privacy-preserving, event-triggered control framework that lets coupled neural networks on signed graphs achieve bipartite synchronization without...

University of Queensland researchers have created PlantMicro, a benchmarking dataset showing that vision language models perform at barely above random...

A new study combines a lightweight MobileNetV2 model with edge computing to detect network attacks with 94.6 percent accuracy in...

Researchers at the Universidad de Salamanca have developed a Lie-algebraic framework that separates task-relevant signal from gradient geometry in variational...

A camera-based AI system running on an ordinary tablet scored the Balance Error Scoring System more consistently than real-time human...

A machine learning model trained on 26 hours of Kitepower flight data can predict critical tether load events one second...

Researchers at the University of Vienna have developed a Bayesian method based on Benford's Law that automatically detects hidden anomalies...