Facebook AI
Paving The Way For Software 2.0 With Kotlin
Our work with differentiable programming, which enables programs to optimize themselves, is part of Facebook AI’s broader efforts to build more advanced tools for machine learning (ML) programming. That’s why we’re extending the Kotlin compiler to make differentiability a first-class feature of the Kotlin language, as well as developing a system for tensor typing. Our …
Facebook AI’s Co-teaching Program To Increase Pathways Into AI For Diverse Candidates
AI is doing tremendous things for our world, like detecting diseases, changing transportation, and helping people connect, but it still has a lot to learn. The more AI learns, the better it will understand the world around us. But equally important is who teaches it: If AI learns from only a small group of people, …
The First AI Model That Translates 100 Languages Without Relying On English Data
Facebook AI is introducing, M2M-100 the first multilingual machine translation (MMT) model that translates between any pair of 100 languages without relying on English data. It’s open sourced here. When translating, say, Chinese to French, previous best multilingual models train on Chinese to English and English to French, because English training data is the most widely available. Our …
Facebook And Carnegie Mellon Launch The Open Catalyst Project To Find New Ways To Store Renewable Energy
Facebook AI and the Carnegie Mellon University (CMU) Department of Chemical Engineering are announcing the Open Catalyst Project, a collaboration intended to use AI to accelerate quantum mechanical simulations by 1,000x in order to discover new electrocatalysts needed for more efficient and scalable ways to store and use renewable energy. Wind and solar energy are …
Using Reinforcement Learning To Personalize AI-accelerated MRI Scans
What the research is: A method leveraging reinforcement learning to improve AI-accelerated magnetic resonance imaging (MRI) scans. Experiments using the fastMRI data set created by NYU Langone show that our models significantly reduce reconstruction errors by dynamically adjusting the sequence of k-space measurements, a process known as active MRI acquisition. These k-space measurements are the building blocks …
Retrieval Augmented Generation: Streamlining The Creation Of Intelligent Natural Language Processing Models
Teaching computers to understand how humans write and speak, known as natural language processing (NLP), is one of the oldest challenges in AI research. There has been a marked change in approach over the past two years, however. Where research once focused on developing specific frameworks for specific tasks, today powerful general-purpose language models can …
Facebook Wants to Make Smart Robots to Explore Every Nook And Cranny Of Your Home
“Hey Alexa, turn on the kitchen light.” “Hey Alexa, play soothing music at volume three.” “Hey Alexa, tell me where to find my keys.” You can ask an Alexa or Google home assistant questions about facts, news, or the weather, and make commands for whatever you’ve synced them to (lights, alarms, TVs, etc.). But helping …
FastMRI Breakthrough Shows AI-accelerated MRIs Interchangeable With Traditional MRIs
Magnetic resonance imaging (MRI) scanners, the machines that doctors use to image the body’s soft tissue and organs, have been crucial tools for medical diagnoses for almost half a century. But one of their major shortcomings is the time it takes to carry out a scan. Clinicians can spend up to an hour gathering sufficient data for a …
PyTorch Drives Next-gen Intelligent Farming Machines
PyTorch is helping to power a new generation of AI-enhanced farming machines. For farmers, weeds pose a very real threat to the health of crops at a time when global population growth is raising food demand while also making resources such as land and water increasingly scarce. Seeking solutions to helping farmers produce more food with fewer …
A Facebook-scale Simulator To Detect Harmful Behaviors
For large-scale social networks, testing a proposed code update or new feature is a complex and challenging task. In person and also online, people act and interact with one another in ways that are sometimes difficult for traditional algorithms to model or replicate. People’s behavior evolves and adapts over time and is different from one …