News
Feature Fields for Robotic Manipulation (F3RM)
By blending 2D images with foundation models to build 3D feature fields, a new MIT method helps robots understand and manipulate nearby objects with open-ended language prompts.
New technique helps robots pack objects into a tight space
Quest researchers are part of an MIT team that coaxed a family of generative AI models to work together to solve multistep robot manipulation problems.
James DiCarlo on Reverse-engineering Human Sight with AI
Professor DiCarlo joins the Theory and Practice podcast
Could neuroscience and AI engineering unlock a hidden visual interface to the emotional brain?
Professor Jim DiCarlo gave a TEDx talk about his group's research on how the brain processes visual stimuli — and how that understanding may lead to improvements in physical and emotional health.
Solving a machine-learning mystery
A new study shows how large language models like GPT-3 can learn a new task from just a few examples, without the need for any new training data.