Category
Research
-
-
Our framework for reporting model misalignment
-
On the Navier–Stokes Millennium Prize Problem
-
Research acceleration: The view inside OpenAI
-
GPT-6 Astra: A new generation of intelligence
-
Solving the solvent problem
-
The benefits of medical AI assistance vary based on user expertise
-
How enabling two settings tripled our scores on the ARC-AGI-3 benchmark
-
Working to automate nuclear plant operations
-
Working to automate nuclear plant operations
-
A better way to turn 2D designs into 3D models for rapid prototyping
-
3 Questions: Neural transparency and the future of AI design
-
Helping AI models to meet the real world
-
Helping AI models to meet the real world
-
AI agents create virtual playgrounds to help robots get crucial training data
-
New method aims to keep kids safe from illegal AI-generated content
-
Tiny robot boats build floating structures
-
Separating signal from noise in coding evaluations
-
How novice coders can develop AI programs for military applications
-
How Open Models Are Driving AI Research
-
MIT in the media: Innovating and educating for the next 250 years of America
-
Introducing GeneBench-Pro
-
LLMs help robots understand vague instructions and focus on key details
-
Improving the speed and energy-efficiency of AI agents
-
New chip could help tiny robots traverse complex environments
-
A better way to model the behavior of metal alloys
-
MIT in the media: For the future of tech, « Massachusetts can absolutely lead »
-
In game theory, generalists sometimes win out over specialists
-
A near-autonomous AI chemist improves a challenging reaction in medicinal chemistry
-
Could AI tell you where you left your keys?
-
Introducing LifeSciBench
-
Predicting model behavior before release by simulating deployment
-
When it comes to predicting people’s preferences, it pays to consider “the power of three”
-
The consequences of relying on AI for accurate news
-
NSF renews support for MIT-led AI and physics institute, expanding a new model for discovery
-
Dreaming: Better memory for a more helpful ChatGPT
-
Teaching AI agents to ask better questions by playing “Battleship”
-
NVIDIA Research Unlocks Advanced Grasping, Smarter Autonomous Driving and Agent Training at Scale
-
MIT researchers teach AI models to interpret charts
-
Media Advisory: MIT to establish regional quantum hub
-
NVIDIA Research Advances Robotics From Simulation to the Real World
-
An OpenAI model has disproved a central conjecture in discrete geometry
-
What Parameter Golf taught us about AI-assisted research
-
Powering the Next American Century: US Energy Secretary Chris Wright and NVIDIA’s Ian Buck on the Genesis Mission
-
Making the case for curiosity-driven science
-
Solving the “Whac-a-mole dilemma”: A smarter way to debias AI vision models
-
The MIT-IBM Computing Research Lab launches to shape the future of AI and quantum computing
-
Enabling privacy-preserving AI training on everyday devices
-
A faster way to estimate AI power consumption
-
Making Sense of the Early Universe
-
Teaching AI models to say “I’m not sure”
-
Introducing OpenAI Privacy Filter
-
Introducing GPT-Rosalind for life sciences research
-
Human-machine teaming dives underwater
-
New technique makes AI models leaner and faster while they’re still learning
-
Helping data centers deliver higher performance with less hardware
-
Evaluating the ethics of autonomous systems
-
Preview tool helps makers visualize 3D-printed objects
-
MIT researchers use AI to uncover atomic defects in materials
-
MIT engineers design proteins by their motion, not just their shape
-
AI system learns to keep warehouse robot traffic running smoothly
-
Augmenting citizen science with computer vision for fish monitoring
-
Wristband enables wearers to control a robotic hand with their own movements
-
Inside our approach to the Model Spec
-
How to create “humble” AI
-
MIT and Hasso Plattner Institute establish collaborative hub for AI and creativity
-
A better method for identifying overconfident large language models
-
Generative AI improves a wireless vision system that sees through obstructions
-
MIT-IBM Watson AI Lab seed to signal: Amplifying early-career faculty impact
-
Can AI help predict which heart-failure patients will worsen within a year?
-
3 Questions: On the future of AI and the mathematical and physical sciences
-
A better method for planning complex visual tasks
-
Improving instruction hierarchy in frontier LLMs
-
Improving AI models’ ability to explain their predictions
-
Reasoning models struggle to control their chains of thought, and that’s good
-
Extending single-minus amplitudes to gravitons
-
A “ChatGPT for spreadsheets” helps solve difficult engineering challenges faster
-
New method could increase LLM training efficiency
-
Mixing generative AI with physics to create personal items that work in the real world
-
AI to help researchers see the bigger picture in cell biology
-
Our First Proof submissions
-
Study: AI chatbots provide less-accurate information to vulnerable users
-
Exposing biases, moods, personalities, and abstract concepts hidden in large language models
-
Parking-aware navigation system could prevent frustration and emissions
-
Personalization features can make LLMs more agreeable
-
Introducing EVMbench
-
GPT-5.2 derives a new result in theoretical physics
-
Accelerating science with AI and simulations
-
Using synthetic biology and AI to address global antimicrobial resistance threat
-
AI algorithm enables tracking of vital white matter pathways
-
3 Questions: Using AI to help Olympic skaters land a quint
-
Study: Platforms that rank the latest LLMs can be unreliable
-
“This is science!” – MIT president talks about the importance of America’s research enterprise on GBH’s Boston Public Radio
-
Helping AI agents search to get the best results out of large language models