Researchers at MIT and the Chinese University of Hong Kong have developed a machine learning model to close the gap between design and manufacturing in the field of photolithography. The technique, which involves manipulating light to etch onto surfaces, sees use in the creation of computer chips and optical devices but often falls short of…
Computational models that mirror the structure and functioning of the human auditory system could lead to improvements in hearing aids, cochlear implants, and brain-machine interfaces, researchers at MIT say. The team has conducted the largest study yet of deep neural networks trained to perform auditory tasks, and found that most generate internal representations bearing similarities…
During a chemical reaction, molecules move towards a transition state, a high-energy state that dictates how the reaction will proceed. However, this transition state is difficult to predict and observe due to its fleeting nature. Traditionally, scientists use quantum chemistry methods like density functional theory to evaluate these transition states, though these calculations tend to…
MIT researchers have developed an approach based on machine learning that can calculate transition states of chemical reactions within seconds. The structures of these transition states, a temporary condition in the middle of a chemical reaction, can typically only be calculated using techniques based on quantum chemistry – a process that can be extremely time-consuming.…
In the field of biomedicine, segmentation refers to the process of highlighting important structures in a medical image, from organs to cells. Artificial intelligence (AI) models are starting to play a pivotal role in this task, but there are limitations with most existing models, mainly due to the fact that they are unable to factor…
A team at MIT, along with the Broad Institute of MIT and Harvard, and Massachusetts General Hospital, has developed an artificial intelligence (AI) tool that can help navigate the uncertainty in medical image analysis. The tool, named Tyche, provides multiple possible interpretations of a medical image rather than the single answer typically provided by AI…
MIT’s Stephen A. Schwarzman College of Computing has opened its new headquarters in Building 45, creating a hub for computing on campus. The building is considered a physical manifestation of the college's mission to fortify core computer science and AI, integrate computing throughout MIT, and advance the social, ethical and policy considerations of the discipline.
MIT…
The MIT Stephen A. Schwarzman College of Computing has recently inaugurated its new headquarters in Building 45, fostering a new hub of connectivity at MIT. The structure serves as a computing crossroads for the campus and aims to catalyze collaborations in computing, and houses research groups from multiple departments and labs.
Approximately 178,000 square feet in…
To improve the planning and problem-solving capabilities of language models, researchers from Stanford University, MIT, and Harvey Mudd have introduced a method called Stream of Search (SoS). This method trains language models on search sequences represented as serialized strings. It essentially presents these models with a set of problems and solutions in the language they…
Language models (LMs) are a crucial segment of artificial intelligence and can play a key role in complex decision-making, planning, and reasoning. However, despite LMs having the capacity to learn and improve, their training often lacks exposure to effective learning from mistakes. Several models also face difficulties in planning and anticipating the consequences of their…
The complexity of mathematical reasoning in large language models (LLMs) often exceed the capabilities of existing evaluation methods. These models are crucial for problem-solving and decision-making, particularly in the field of artificial intelligence (AI). Yet the primary method of evaluation – comparing the final LLM result to a ground truth and then calculating overall accuracy…
Generative language models in the field of natural language processing (NLP) have fuelled significant progression, largely due to the availability of a vast amount of web-scale textual data. Such models can analyze and learn complex linguistic structures and patterns, which are subsequently used for various tasks. However, successful implementation of these models depends heavily on…