MIT and Google researchers discover how GPT-3 can learn new tasks with minimal data

367
Scientists from the Massachusetts Institute of Technology (MIT), Google Research and Stanford University have published a new study which shows how large language models, like GPT-3, can learn new tasks, using a small number of examples and without attracting new training data. According to the researchers, neural network models can contain smaller and simpler linear models within them. This means that a simple training algorithm can be used in a large model that will train a linear model to solve a particular task, using only the information already contained in the large model. If the results of this study are confirmed, it will be an important step toward understanding how contextual learning works.

Similar research is being conducted by a team from the same Massachusetts Institute of Technology and the joint MIT and IBM Watson Artificial Intelligence Laboratory. They have developed a technique that allows the model to quantify uncertainty more efficiently with less computing power than is used in other techniques. It also does not require additional data. This technique does not involve additional retraining or model modification, and can be used in most applications. It includes a simple complementary model that helps the basic ML model calculate uncertainty.


18 countries have unveiled the first international agreement on how to protect artificial intelligence from irresponsible players. It aims to develop AI solutions that are "inherently safe".

On November 30, the professional IT community GlobalCIO hosted a large-scaled international conference "Global CIO Insights: Digital Transformation with AI". During the event, leading experts shared their practical experience in launching projects utilizing artificial intelligence (AI) and highlighted approaches that helped elevate their companies to new heights.

Voting for projects participating in the "Project of the Year" contest is open. The voting began on December 1st and will continue until January 15th inclusive. The winners will be announced on February 7th, 2024.

Online sales is one of the areas where the quality of IT tools directly affects business profitability. Kamza Nugumanov, CIO of Jusan store, tells about the experience of deploying a rapidly growing Kazakh marketplace.

We use cookies for analytical purposes and to deliver you the best experience with our website. Continuing to the site, you agree to the Cookie Policy.