Research Focus
Meta's AI research organization (formerly known as FAIR, Facebook AI Research) has historically emphasized open publication and released tools, spanning computer vision, self-supervised learning, and the Llama family of language models.
Notable Contributions
- PyTorch, a widely-adopted open-source deep learning framework originating from Meta
- The Llama model series, released with open weights
- Segment Anything (SAM), an influential open computer vision model
- Substantial published work on self-supervised learning methods
Staying Current on This Lab's Work
See our Meta news for recent product announcements, or the lab's own official research blog for primary publications.
Related Pages
Frequently Asked
Is Meta's AI research typically open or closed?
Historically emphasized open publication and open-weight releases, though the company's strategy has been evolving, with some more recent work developed under tighter control.
What is PyTorch and where did it come from?
A widely-used open-source deep learning framework that originated from Meta's AI research organization.
Where can I read Meta's official research?
Check Meta AI's official research page for primary publications.
Where can I see Meta's specific model papers?
See our Llama 4 page.