MIT researchers develop an efficient way to train more reliable AI agents
The technique could make AI systems better at complex tasks that involve variability.
National Science Foundation (NSF)
Despite its impressive output, generative AI doesn’t have a coherent understanding of the world
Researchers show that even the best-performing large language models don’t form a true model of the world and its rules, and can thus fail unexpectedly on similar tasks.
Enhancing LLM collaboration for smarter, more efficient solutions
“Co-LLM” algorithm helps a general-purpose AI model collaborate with an expert large language model by combining the best parts of both answers, leading to more factual responses.
Helping robots practice skills independently to adapt to unfamiliar environments
New algorithm helps robots practice skills like sweeping and placing objects, potentially helping them improve at important tasks in houses, hospitals, and factories.
MIT researchers advance automated interpretability in AI models
MAIA is a multimodal agent that can iteratively design experiments to better understand various components of AI systems.
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.
A smarter way to streamline drug discovery
The SPARROW algorithm automatically identifies the best molecules to test as potential new medicines, given the vast number of factors affecting each choice.
A technique for more effective multipurpose robots
With generative AI models, researchers combined robotics data from different sources to help robots learn better.
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.
This tiny chip can safeguard user data while enabling efficient computing on a smartphone
Researchers have developed a security solution for power-hungry AI models that offers protection against two common attacks.