A new way to test how well AI systems classify text
As large language models increasingly dominate our everyday lives, new systems for checking their reliability are more important than ever.
As large language models increasingly dominate our everyday lives, new systems for checking their reliability are more important than ever.
Training an ever-growing percentage of MIT’s students, the Department of Electrical Engineering and Computer Science relies heavily on dedicated and passionate TAs.
New research can identify opportunities to drive down the cost of renewable energy systems, batteries, and many other technologies.
Ianacare, co-founded by Steven Lee ’97, MEng ’98, equips caregivers with the resources, networks, and tools they need to support loved ones.
New research shows automatically controlling vehicle speeds to mitigate traffic at intersections can cut carbon emissions between 11 and 22 percent.
By visualizing Escher-like optical illusions in 2.5 dimensions, the “Meschers” tool could help scientists understand physics-defying shapes and spark new designs.
AeroAstro professor and outgoing co-director of the Center for Computational Science and Engineering will play a vital role in fostering community for bilingual computing faculty.
New professors join Comparative Media Studies/Writing, History, Linguistics and Philosophy, Music and Theater Arts, and Political Science.
This new approach could lead to enhanced AI models for drug and materials discovery.
The flexible chip could boost the performance of current electronics and meet the more stringent efficiency requirements of future 6G technologies.
The platform identifies, mixes, and tests up to 700 new polymer blends a day for applications like protein stabilization, battery electrolytes, or drug-delivery materials.
Neural Jacobian Fields, developed by MIT CSAIL researchers, can learn to control any robot from a single camera, without any other sensors.
An oft-ignored effect can be used to probe an important property of semiconductors, a new study finds.
MIT researchers found that special kinds of neural networks, called encoders or “tokenizers,” can do much more than previously realized.
Language models follow changing situations using clever arithmetic, instead of sequential tracking. By controlling when these approaches are used, engineers could improve the systems’ capabilities.