Language Modelling at Scale: Gopher, Ethical considerations, and Retrieval
We are releasing three papers on language models, Gopher, Ethical considerations, and Retrieval. Read More
SaiSatish
Engineers teach AI to navigate ocean with minimal energy
New AI uses reinforcement learning to efficiently navigate oceans Read More
These tiny liquid robots never run out of juice as long as they have food
Scientists at Berkeley Lab and the University of Massachusetts Amherst have demonstrated the first self-powered, aqueous robot that runs continuously without electricity. The technology has potential as an automated chemical synthesis or drug delivery system for pharmaceuticals. Read More
Improving Vision Transformer Efficiency and Accuracy by Learning to Tokenize
Posted by Michael Ryoo, Research Scientist, Robotics at Google and Anurag Arnab, Research Scientist, Google Research Transformer models consistently obtain state-of-the-art results in computer vision tasks, including object detection and video classification. In contrast to standard convolutional approaches that process…
Google at NeurIPS 2021
Posted by Jaqui Herman and Cat Armato, Program Managers This week marks the beginning of the 35th annual Conference on Neural Information Processing Systems (NeurIPS 2021), the biggest machine learning conference of the year. NeurIPS 2021 will be held virtually…
Evaluating Syntactic Abilities of Language Models
Posted by Jason Wei, AI Resident and Dan Garrette, Research Scientist, Google Research In recent years, pre-trained language models, such as BERT and GPT-3, have seen widespread use in natural language processing (NLP). By training on large volumes of text,…
How statistics can aid in the fight against misinformation
Mathematicians created a statistical model that can be used to detect misinformation in social posts. Read More
Light-powered soft robots could suck up oil spills
A floating, robotic film could be trained to hoover oil spills at sea or remove contaminants from drinking water. Read More
RLDS: An Ecosystem to Generate, Share, and Use Datasets in Reinforcement Learning
Posted by Sabela Ramos, Software Engineer and Léonard Hussenot, Student Researcher, Google Research, Brain Team Most reinforcement learning (RL) and sequential decision making algorithms require an agent to generate training data through large amounts of interactions with their environment to…
‘My robot is a softie’: Physical texture influences judgments of robot personality
Researchers have found that the physical texture of robots influenced perceptions of robot personality. Furthermore, first impressions of robots, based on physical appearance alone, could influence the relationship between physical texture and robot personality formation. This work could facilitate the…