Add BrightSurf on Google Email

Massachusetts Institute of Technology


Deep learning with light

Researchers at MIT have developed a new method that uses optics to accelerate machine-learning computations on low-power devices. By encoding model components onto light waves, data can be transmitted rapidly and computations performed quickly, leading to over a hundredfold improvement in energy efficiency.

Method for decoding asteroid interiors could help aim asteroid-deflecting missions

Researchers have developed a method to map an asteroid's interior structure based on its spin changes during close encounters with Earth. This technique, called AIME, could help scientists plan more effective defense strategies for asteroids like Apophis, which poses a significant hazard if it were to make impact.

SourceMassachusetts Institute of Technology·JournalMonthly Notices of the Royal Astronomical Society·DateOct 19, 2022

What drives ecosystems to instability?

Researchers have developed a phase diagram that predicts ecosystem stability based on the number of species and interaction strength. The study shows three states of ecological communities, with conditions for moving between them, allowing for the prediction of instability and fluctuations in populations. This work builds on previous f...

Learning on the edge

Researchers developed a new technique that enables on-device training using less than a quarter of a megabyte of memory, reducing the need for powerful computers and central servers. This approach preserves privacy by keeping data on the device, making deep learning more accessible for low-power edge devices.

Study: Astronomers risk misinterpreting planetary signals in James Webb data

A new MIT study suggests that current opacity models used by astronomers may not be accurate enough to interpret the precise light-based signals from the James Webb Space Telescope. The researchers predict that properties of planetary atmospheres, such as temperature and elemental composition, could be off by an order of magnitude if e...

SourceMassachusetts Institute of Technology·JournalNature Astronomy·DateSep 15, 2022

These neurons have food on the brain

A study from MIT neuroscientists has identified a population of neurons in the visual cortex that respond to images of food. The researchers found four previously known populations and a fifth, more surprising population that appears to be selective for food images. This finding may reflect the special importance of food in human culture.

SourceMassachusetts Institute of Technology·JournalCurrent Biology·DateAug 25, 2022