Using language to give robots a better grasp of an open-ended world
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.
Topic
Computer Science and Artificial Intelligence Laboratory (CSAIL)
Download RSS feed: News Articles / In the Media / Audio
New techniques efficiently accelerate sparse tensors for massive AI models
Complimentary approaches — “HighLight” and “Tailors and Swiftiles” — could boost the performance of demanding machine-learning tasks.
Accelerating AI tasks while preserving data security
The SecureLoop search tool efficiently identifies secure designs for hardware that can boost the performance of complex AI tasks, while requiring less energy.
Making genetic prediction models more inclusive
MIT computer scientists developed a way to calculate polygenic scores that makes them more accurate for people across diverse ancestries.
A new way to integrate data with physical objects
StructCode, developed by MIT CSAIL researchers, encodes machine-readable data in laser-cut objects by modifying their fabrication features.
New technique helps robots pack objects into a tight space
Researchers coaxed a family of generative AI models to work together to solve multistep robot manipulation problems.
Twelve with MIT ties elected to the National Academy of Medicine for 2023
Five MIT faculty, along with seven additional affiliates, are honored for outstanding contributions to medical research.
Empowering students to bring change in the Middle East
The Middle East Entrepreneurs of Tomorrow (MEET) program uses an MIT-inspired curriculum and MISTI student instructors to help young Palestinians and Israelis find common ground.
Finger-shaped sensor enables more dexterous robots
MIT engineers develop a long, curved touch sensor that could enable a robot to grasp and manipulate objects in multiple ways.
AI copilot enhances human precision for safer aviation
Designed to ensure safer skies, “Air-Guardian” blends human intuition with machine precision, creating a more symbiotic relationship between pilot and aircraft.
Improving accessibility of online graphics for blind users
Open-source software by MIT MAD Fellow Jonathan Zong and others in the MIT Visualization Group reveals online graphics’ embedded data in the user’s preferred degree of granularity.
Decoding the complexity of Alzheimer’s disease
By analyzing epigenomic and gene expression changes that occur in Alzheimer’s disease, researchers identify cellular pathways that could become new drug targets.
From physics to generative AI: An AI model for advanced pattern generation
Inspired by physics, a new generative model PFGM++ outperforms diffusion models in image generation.
Five MIT faculty members named 2023 Simons Investigators
The program supports “outstanding theoretical scientists.”
Multi-AI collaboration helps reasoning and factual accuracy in large language models
Researchers use multiple AI models to collaborate, debate, and improve their reasoning abilities to advance the performance of LLMs while increasing accountability and factual accuracy.