Enabling small language models to solve complex reasoning tasks
The “self-steering” DisCIPL system directs small models to work together on tasks with constraints, like itinerary planning and budgeting.
Topic
MIT Siegel Family Quest for Intelligence
Download RSS feed: News Articles / In the Media / Audio
MIT tool visualizes and edits “physically impossible” objects
By visualizing Escher-like optical illusions in 2.5 dimensions, the “Meschers” tool could help scientists understand physics-defying shapes and spark new designs.
The unique, mathematical shortcuts language models use to predict dynamic scenarios
Language models follow changing situations using clever arithmetic, instead of sequential tracking. By controlling when these approaches are used, engineers could improve the systems’ capabilities.
New system enables robots to solve manipulation problems in seconds
Researchers developed an algorithm that lets a robot “think ahead” and consider thousands of potential motion plans simultaneously.
A visual pathway in the brain may do more than recognize objects
New research using computational vision models suggests the brain’s “ventral stream” might be more versatile than previously thought.
Teaching a robot its limits, to complete open-ended tasks safely
The “PRoC3S” method helps an LLM create a viable action plan by testing each step in a simulation. This strategy could eventually aid in-home robots to complete more ambiguous chore requests.
3 Questions: Claire Wang on training the brain for memory sports
The MIT sophomore and award-winning memory champion explains what these competitions are all about and why you might want to build a “memory palace.”
Making it easier to verify an AI model’s responses
By allowing users to clearly see data referenced by a large language model, this tool speeds manual validation to help users spot AI errors.
Helping robots practice skills independently to adapt to unfamiliar environments
A new algorithm helps robots practice skills like sweeping and placing objects, potentially helping them improve at important tasks in houses, hospitals, and factories.
Reasoning skills of large language models are often overestimated
New CSAIL research highlights how LLMs excel in familiar scenarios but struggle in novel ones, questioning their true reasoning abilities versus reliance on memorization.
Natural language boosts LLM performance in coding, planning, and robotics
Three neurosymbolic methods help language models find better abstractions within natural language, then use those representations to execute complex tasks.
A crossroads for computing at MIT
The MIT Schwarzman College of Computing building will form a new cluster of connectivity across a spectrum of disciplines in computing and artificial intelligence.
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.
When computer vision works more like a brain, it sees more like people do
Training artificial neural networks with data from real brains can make computer vision more robust.
Automating the math for decision-making under uncertainty
A new tool brings the benefits of AI programming to a much broader class of problems.