MIT affiliates win AI for Math grants to accelerate mathematical discovery
Department of Mathematics researchers David Roe and Andrew Sutherland seek to advance automated theorem proving; four additional MIT alumni also awarded.
Data
How to build AI scaling laws for efficient LLM training and budget maximization
MIT-IBM Watson AI Lab researchers have developed a universal guide for estimating how large language models will perform based on smaller models in the same family.
3 Questions: The pros and cons of synthetic data in AI
Artificially created data offer benefits from cost savings to privacy preservation, but their limitations require careful planning and evaluation, Kalyan Veeramachaneni says.
3 Questions: On biology and medicine’s “data revolution”
Professor Caroline Uhler discusses her work at the Schmidt Center, thorny problems in math, and the ongoing quest to understand some of the most complex interactions in biology.
MIT researchers develop AI tool to improve flu vaccine strain selection
VaxSeer uses machine learning to predict virus evolution and antigenicity, aiming to make vaccine selection more accurate and less reliant on guesswork.
Simpler models can outperform deep learning at climate prediction
New research shows the natural variability in climate data can cause AI models to struggle at predicting local temperature and rainfall.
Helping data storage keep up with the AI revolution
Storage systems from Cloudian, co-founded by an MIT alumnus, are helping businesses feed data-hungry AI models and agents at scale.
New algorithms enable efficient machine learning with symmetric data
This new approach could lead to enhanced AI models for drug and materials discovery.
The unique, mathematical shortcuts language models use to predict dynamic scenarios
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.
Study could lead to LLMs that are better at complex reasoning
Researchers developed a way to make large language models more adaptable to challenging tasks like strategic planning or process optimization.