Breaking AIs to make them better

Current AIs are very accurate but inflexible at image recognition. Exactly why this is remains a mystery. Researchers have developed a method called 'Raw Zero-Shot' to assess how neural networks handle elements unknown to them. The results have the potential to help researchers identify the common features that make neural networks 'non-robust,' and develop methods to make AIs more reliable.

from Latest Science News -- ScienceDaily https://ift.tt/p8EHwU3

Comments

Post a Comment

Popular posts from this blog

Seismic study reveals key reason why Patagonia is rising as glaciers melt

The icefields that stretch for hundreds of miles atop the Andes mountain range in Chile and Argentina are melting at some of the fastest rates on the planet. The ground that was beneath this ice is also shifting and rising as these glaciers disappear. Geologists have discovered a link between recent ice mass loss, rapid rock uplift and a gap between tectonic plates that underlie Patagonia. from Phys.org - latest science and technology news stories https://ift.tt/j1h2ds5
Read more

New method to detect and visualize sperm cells recovered from forensic evidence

One of the most valuable forms of forensic evidence in cases of male-perpetrated sexual assault is the presence of semen, either in the form of stains left behind on items or on swabs collected from victims after an offense. To confirm that semen is present, suspected stains are examined under a microscope to see if any sperm cells are visible. from Phys.org - latest science and technology news stories https://ift.tt/2XBwacB
Read more