What is the news?
Researchers have developed a highly efficient artificial intelligence (AI) model by mimicking the human brain’s visual system, providing clues to how living brains are able to do so much with so little energy. This unprecedented research has been published in the journal ‘Nature’. The team started with a model with 60 million variables, but managed to compress it into a nearly identical performing version with just 10,000 variables.
Will help in detecting Alzheimer’s disease
“It’s incredibly small. We can send it via Tweet or email,” says study author Ben Cowley, an assistant professor at Cold Spring Harbor Laboratory. This tiny AI model not only works efficiently but also mimics the functioning of a living brain. This opens up new avenues for studying neurological disorders such as Alzheimer’s disease. This study also contributes to the understanding of the human visual system.
Training
Model trained with monkey data
The researchers’ goal was to compress large models into a smaller and more compact form. They started with a model trained on data from macaque monkeys and searched for redundant parts using statistical techniques used to compress digital photos. The result was a small model that could be sent as an email attachment. This gave the team the opportunity to observe the activities of artificial neurons.
Follow Us On Google News
Follow Us On WhatsApp
Leave a Comment