This new machine-learning model can match corresponding audio and visual data, which could someday help robots interact in the real world.
Three neurosymbolic methods help language models find better abstractions within natural language, then use those representations to execute complex tasks.
The Toyota Research Institute (TRI) announced today that it will be extending its AI research collaboration with MIT. TRI, which also has existing relationships with Stanford and the University of Michigan, has selected 13 additional academic institutions to participate in the next five-year phase of its research initiative.
New research reveals biases in fake news datasets and improves the use of automatic detectors.
The Tree-D Fusion system integrates generative AI and genus-conditioned algorithms to create precise simulation-ready models of 600,000 existing urban trees across North America.
Starting with a single frame in a simulation, a new system uses generative AI to emulate the dynamics of molecules, connecting static molecular structures and developing blurry pictures into videos.
Neural Decipherment research named Netexplo/UNESCO award winner
The approach maintains an AI model’s accuracy while ensuring attackers can’t extract secret information.
Scientists came up with a learning pipeline for the MIT Mini-Cheetah that learns to run entirely by trial and error in simulation