Reasoning skills of large language models are often overestimated
New CSAIL research highlights how LLMs excel in familiar scenarios but struggle in novel ones, questioning their true reasoning abilities versus reliance on memorization.
Search Stories
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.
MIT ARCLab announces winners of inaugural Prize for AI Innovation in Space
The challenge asked teams to develop AI algorithms to track and predict satellites’ patterns of life in orbit using passively collected data
Empowering future innovators through a social impact lens
The IDEAS Social Innovation Challenge helps students hone their entrepreneurship skills to create viable ventures for public good.
“They can see themselves shaping the world they live in”
Developed by MIT RAISE, the Day of AI curriculum empowers K-12 students to collaborate on local and global challenges using AI.
MIT researchers introduce generative AI for databases
This new tool offers an easier way for people to analyze complex tabular data.
Study reveals why AI models that analyze medical images can be biased
These models, which can predict a patient’s race, gender, and age, seem to use those traits as shortcuts when making medical diagnoses.
Melissa Choi named director of MIT Lincoln Laboratory
With decades of experience working across the laboratory’s R&D areas, Choi brings a focus on collaboration, technical excellence, and unity.
Startup aims to transform the power grid with superconducting transmission lines
VEIR, founded by alumnus Tim Heidel, has developed technology that can move more power over long distances, with the same footprint as traditional lines.
Designing for outer space
With NASA planning permanent bases in space and on the moon, MIT students develop prototypes for habitats far from planet Earth.
Helping nonexperts build advanced generative AI models
MosaicML, co-founded by an MIT alumnus and a professor, made deep-learning models faster and more efficient. Its acquisition by Databricks broadened that mission.
3 Questions: Catherine D’Ignazio on data science and a quest for justice
In her new book, “Counting Feminicide,” D’Ignazio explores how activists have tabulated the epidemic of gender-based murder in Latin America.
MIT-Takeda Program wraps up with 16 publications, a patent, and nearly two dozen projects completed
The program focused on AI in health care, drawing on Takeda’s R&D experience in drug development and MIT’s deep expertise in AI.
Researchers leverage shadows to model 3D scenes, including objects blocked from view
This technique could lead to safer autonomous vehicles, more efficient AR/VR headsets, or faster warehouse robots.