A single-core-neuron framework unlocks efficient deep learning
Peer-Reviewed Publication
Updates every hour. Last Updated: 18-May-2025 07:10 ET (18-May-2025 11:10 GMT/UTC)
In a paper published in National Science Review, an international team of researchers introduced the one-core-neuron system (OCNS), a “small model” framework designed to tackle the inefficiency of “large models”. By utilizing a single core neuron, OCNS achieves comparable or superior performance in time-series forecasting while reducing parameter requirements to a fraction of those in existing “large models”. This approach highlights the potential of small models in efficient deep learning and more sustainable AI applications.
HONG KONG (8 January 2025) — Recognised for his leading position in digital medicine, Professor Yu Xinge from the Department of Biomedical Engineering at City University of Hong Kong (CityUHK) was honoured with the 3rd Zhong Nanshan Youth Science and Technology Innovation Award. This prestigious accolade recognises exceptional young scientists in medicine and life sciences.
In research published in Advanced Intelligence Systems, scientists have developed an innovative, soft, wearable robot to help workers avoid job-related injuries while lifting, lowering, and carrying objects.
Genome editing with CRISPR-Cas is often associated with the induction of mutations. However, a team of researchers from the Swiss University of Lausanne now shows that it can also be used to repair natural mutations.