Meta science summit expands brain data goal and releases UMA Nitro
Original title: Advancing Science Together: Highlights From the Open Source AI For Science Summit
FAIR raised the brain-data target from 15,000 to 25,000 hours and showed UMA finding drug crystal forms without extra training.
Meta’s FAIR team hosted the Open Source AI for Science Summit in Menlo Park and recapped partnerships in healthcare, chemistry, and mathematics. It expanded the Digital Brain Project with the Rothschild Hospital Foundation and the University of Montreal. Meta released UMA Nitro, a faster Universal Model for Atoms, and described classroom use of UMA Playground plus the open ATLAS math library.
- The Digital Brain Project data target rose from 15,000 hours to 25,000.
- A Carnegie Mellon and UC Riverside study used UMA models on drug crystal forms.
- ATLAS formalizes 26 open-access textbooks and more than 45,000 checked results.
- Discovery Education says it reaches more than 4 million educators and 45 million students.
- University of Pittsburgh showed assistive robotics built on DINOv3 and SAM 3.
From the original article
3 min read at about.fb.comWe brought researchers from universities, hospitals, and industry together for the Open Source AI for Science Summit, an event built around using open source AI to accelerate scientific discovery. The post Advancing Science Together: Highlights From the Open Source AI For Science Summit appeared first on Meta Newsroom.
By Facebook · about.fb.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.