New insights into training dynamics of deep classifiers
MIT researchers uncover the structural properties and dynamics of deep classifiers, offering novel explanations for optimization, generalization, and approximation in deep networks.
Large language models are biased. Can logic help save them?
MIT researchers trained logic-aware language models to reduce harmful stereotypes like gender and racial biases.
Robot armies duke it out in Battlecode’s epic on-screen battles
The long-running programming competition encourages skills and friendships that last a lifetime.
Phiala Shanahan is seeking fundamental answers about our physical world
With supercomputers and machine learning, the physicist aims to illuminate the structure of everyday particles and uncover signs of dark matter.
MIT-Takeda Program heads into fourth year with crop of 10 new projects
The program leverages MIT’s research expertise and Takeda’s industrial know-how for research in artificial intelligence and medicine.
Improving the speed and safety of airport security screening
Lincoln Laboratory seeks ways to build non-contact screening methods that can detect concealed explosives at airports.
Efficient technique improves machine-learning models’ reliability
The method enables a model to determine its confidence in a prediction, while using no additional data and far fewer computing resources than other methods.
Helping companies deploy AI models more responsibly
MIT spinout Verta offers tools to help companies introduce, monitor, and manage machine-learning models safely and at scale.
Solving a machine-learning mystery
A new study shows how large language models like GPT-3 can learn a new task from just a few examples, without the need for any new training data.
Automating the math for decision-making under uncertainty
A new tool brings the benefits of AI programming to a much broader class of problems.
How to make hydrogels more injectable
A new computational framework could help researchers design granular hydrogels to repair or replace diseased tissues.
Putting clear bounds on uncertainty
Computer scientists want to know the exact limits in our ability to clean up, and reconstruct, partly blurred images.
Research, education, and connection in the face of war
“I wouldn’t let the aggressor in the war squash my dreams,” says Ukrainian mathematician and MITx MicroMasters learner Tetiana Herasymova.
MIT researchers develop an AI model that can detect future lung cancer risk
Deep-learning model takes a personalized approach to assessing each patient’s risk of lung cancer based on CT scans.
Gaining real-world industry experience through Break Through Tech AI at MIT
A new experiential learning opportunity challenges undergraduates across the Greater Boston area to apply their AI skills to a range of industry projects.