Unpacking black-box models
Researchers create a mathematical framework to evaluate explanations of machine-learning models and quantify how well people understand them.
Search Stories
Artificial intelligence system learns concepts shared across video, audio, and text
A machine-learning model can identify the action in a video clip and label it, without the help of humans.
What words can convey
Natural language processing models capture rich knowledge of words’ meanings through statistics.
Sanjay Sarma to step down as vice president for open learning
Following nine years leading MIT’s work on digital learning, the longtime faculty member will focus on teaching, research, writing, and entrepreneurship.
Soft assistive robotic wearables get a boost from rapid design tool
Scientists have created a design and fabrication tool for soft pneumatic actuators for integrated sensing, which can power personalized health care, smart homes, and gaming.
Secret superheroes of EECS
Lecturers ensure the technology, curriculum, and instructional delivery of MIT’s classroom education keep up with the dizzying pace of its research.
A one-up on motion capture
A new neural network approach captures the characteristics of a physical system’s dynamic motion from video, regardless of rendering configuration or image differences.
Engineers use artificial intelligence to capture the complexity of breaking waves
Their model’s predictions should help researchers improve ocean climate simulations and hone the design of offshore structures.
Surface coating designed to improve power plant efficiency wins 2022 Water Innovation Prize
The student pitch competition included a variety of solutions addressing water access, usage, and maintenance.
Aging Brain Initiative awards fund five new ideas to study, fight neurodegeneration
Competitive seed grants launch yearlong investigations of novel hypotheses about potential causes, biomarkers, treatments of Alzheimer’s and ALS.
Multiplying the MIT $100K’s impact
The $100K Entrepreneurship Competition has inspired students to bring the model to regions across the globe.
3 Questions: Designing software for research ethics
PhD candidate Jonathan Zong found a lack of systems that earn and maintain public trust in large-scale online research — so he made one himself.
A smarter way to develop new drugs
A new artificial intelligence technique only proposes candidate molecules that can actually be produced in a lab.
Estimating the informativeness of data
MIT researchers can now estimate how much information data are likely to contain, in a more accurate and scalable way than previous methods.
An easier way to teach robots new skills
Researchers have developed a technique that enables a robot to learn a new pick-and-place task with only a handful of human demonstrations.