Training LLMs to self-detoxify their language
A new method from the MIT-IBM Watson AI Lab helps large language models to steer their own responses toward safer, more ethical, value-aligned outputs.
A new method from the MIT-IBM Watson AI Lab helps large language models to steer their own responses toward safer, more ethical, value-aligned outputs.
With projected global warming, the frequency of extreme storms will ramp up by the end of the century, according to a new study.
The Hood Pediatric Innovation Hub aims to break down barriers to pediatric innovation and foster transformative research to improve children’s health outcomes.
These bacteria, which could be designed to detect pollution or nutrients, could act as sensors to help farmers monitor their crops.
The approach maintains an AI model’s accuracy while ensuring attackers can’t extract secret information.
New research finds liquid air energy storage could be the lowest-cost option for ensuring a continuous power supply on a future grid dominated by carbon-free but intermittent sources of electricity.
MIT engineers developed an insect-sized jumping robot that can traverse challenging terrains and carry heavy payloads.
A new method lets users ask, in plain language, for a new molecule with certain properties, and receive a detailed description of how to synthesize it.
Researchers analyzed the full lifecycle of several fuel options and found this approach has a comparable environmental impact, overall, to burning low-sulfur fuels.
Political scientist Kathleen Thelen’s new book explains how America’s large retailers got very, very large.
“InteRecon” enables users to capture items in a mobile app and reconstruct their interactive features in mixed reality. The tool could assist in education, medical environments, museums, and more.
Connected by the MIT Human Insight Collaborative, Lecturer Mi-Eun Kim and Research Scientist Praneeth Namburi want to develop an understanding of musical expression and skill development.
New research on a cytokine called IL-17 adds to growing evidence that immune molecules can influence behavior during illness.
The framework helps clinicians choose phrases that more accurately reflect the likelihood that certain conditions are present in X-rays.
Upending a long-held supposition, MIT researchers find a common catalyst works by cycling between two different forms.