Improving the speed and energy-efficiency of AI agents
A new system, known as Murakkab, optimizes the design and deployment of multistep workflows that power AI applications.
In game theory, generalists sometimes win out over specialists
Researchers show that for certain kinds of games, an overlooked class of algorithms performs much better than expected.
Powering 160,000 hours of discovery at MIT.nano
NanoFab Equipment Management and Operations (NEMO) system streamlines shared facilities management via tool trainings, reservations, and lab communications.
Professor Emeritus Jack Dennis, pioneering developer of dataflow models of computation, dies at 94
The influential first leader of the Computation Structures Group at MIT played a key role in the development of asynchronous computing.
Bridging space research and policy
PhD student Carissma McGee studies exoplanets and examines intellectual property frameworks for space collaborations.
Engineering confidence to navigate uncertainty
In 16.85 (Design and Testing of Autonomous Vehicles), AeroAstro students build software that allows autonomous flight vehicles to navigate unknown environments.
Exposing biases, moods, personalities, and abstract concepts hidden in large language models
A new method developed at MIT could root out vulnerabilities and improve LLM safety and performance.
Helping AI agents search to get the best results out of large language models
EnCompass executes AI agent programs by backtracking and making multiple attempts, finding the best set of outputs generated by an LLM. It could help coders work with AI agents more efficiently.
Helping companies with physical operations around the world run more intelligently
Founded by two MIT alumni, Samsara’s platform gives companies a central hub to learn from their workers, equipment, and other infrastructure.
MIT Sea Grant students explore the intersection of technology and offshore aquaculture in Norway
AquaCulture Shock program, in collaboration with MIT-Scandinavia MISTI, offers international internships for AI and autonomy in aquaculture
New AI agent learns to use CAD to create 3D objects from sketches
The virtual VideoCAD tool could boost designers’ productivity and help train engineers learning computer-aided design.
MIT researchers propose a new model for legible, modular software
The coding framework uses modular concepts and simple synchronization rules to make software clearer, safer, and easier for LLMs to generate.
New machine-learning application to help researchers predict chemical properties
ChemXploreML makes advanced chemical predictions easier and faster — without requiring deep programming skills.
Can AI really code? Study maps the roadblocks to autonomous software engineering
A team of researchers has mapped the challenges of AI in software development, and outlined a research agenda to move the field forward.
Memory safety is at a tipping point
Lincoln Laboratory cybersecurity expert Hamed Okhravi calls for a unified approach to securing computer memory, as a matter of national security.