When to trust an AI model
More accurate uncertainty estimates could help users decide about how and when to use machine-learning models in the real world.
More accurate uncertainty estimates could help users decide about how and when to use machine-learning models in the real world.
Ammonia could be a nearly carbon-free maritime fuel, but without new emissions regulations, its impact on air quality could significantly impact human health.
While women and men self-reported similar vaccination rates, unvaccinated women had less intention to get vaccinated than men.
A chip the size of a pack of cards uses fewer resources and a smaller footprint than existing automated manufacturing platforms and could lead to more affordable cell therapy manufacturing.
A study by MIT scientists supports “social good” as a cognitive approach to dealing with highly stressful events.
New findings challenge current thinking on the ocean’s role in storing carbon.
This new tool offers an easier way for people to analyze complex tabular data.
By helping microbes withstand industrial processing, the method could make it easier to harness the benefits of microorganisms used as medicines and in agriculture.
A national bottle deposit fee could make a dramatic difference in reducing plastic waste, MIT researchers report.
Rising superpowers like China are “cautious opportunists” in global institutions, and the U.S. should avoid overreaction, PhD student Raymond Wang argues.
The new design approach could be used to produce metals with exceptional combinations of strength and ductility, for aerospace and other applications.
This tiny, biocompatible sensor may overcome one of the biggest hurdles that prevent the devices from being completely implanted.
A new surgical procedure gives people more neural feedback from their residual limb. With it, seven patients walked more naturally and navigated obstacles.
The newly synthesized material could be the basis for wearable thermoelectric and spintronic devices.
These models, which can predict a patient’s race, gender, and age, seem to use those traits as shortcuts when making medical diagnoses.