DeepMind   @DeepMind

Our team research and build safe AI systems. We're committed to solving intelligence, to advance science and humanity.





  Tweets by DeepMind  

DeepMind    @DeepMind   ·   10/18/2021
We’ve acquired the MuJoCo physics simulator (https://t.co/knwXLZMr4L) and are making it free for all, to support research everywhere. MuJoCo is a fast, powerful, easy-to-use, and soon to be open-source simulation tool, designed for robotics research: https://t.co/Of3Q1W2GIR
 
 Reply      Retweet   644      Like     2526    

More by:   DeepMind
zoom in  
 



DeepMind    @DeepMind   ·   7/6/2021
Many models bake in domain knowledge to control how input data is processed. This means models must be redesigned to handle new types of data. Introducing the Perceiver, an architecture that works on many kinds of data - in some cases all at once: https://t.co/2OAMwt61ru (1/)
 Reply      Retweet   241      Like     941    

More by:   DeepMind
zoom in  
 



DeepMind    @DeepMind   ·   6/16/2020
Moving away from negative pairs in self-supervised representation learning: our new SotA method, Bootstrap Your Own Latent (BYOL), narrows the gap between self-supervised & supervised methods simply by predicting previous versions of itself. See here: https://t.co/qyaSXnPQjN
 
 Reply      Retweet   238      Like     837    

More by:   DeepMind
zoom in  
 



DeepMind    @DeepMind   ·   7/16/2021
Social interactions are key to intelligence, but do artificial agents understand this? Introducing Melting Pot, an evaluation suite for reinforcement learning agents that tests their socio-cognitive skills: https://t.co/FhUqyqZODf OS: https://t.co/AFa4H6peQn #ICML2021
 Reply      Retweet   113      Like     370    

More by:   DeepMind
zoom in  
 



DeepMind    @DeepMind   ·   10/4/2021
Published today in @naturemethods together with colleagues from @calico: Enformer - a transformer model that has led to greatly increased accuracy in predicting gene expression from DNA sequence. Blog: https://t.co/21zIewaiKR Paper: https://t.co/Bg1gPp4AWe 1/
 
 Reply      Retweet   90      Like     326    

More by:   DeepMind
zoom in  
 



DeepMind    @DeepMind   ·   10/28/2020
Bootstrap Your Own Latent (BYOL) learns useful representations without explicit negative examples. Our researchers show that it also works without using batch statistics at all, challenging the recent "implicit negatives from BatchNorm" hypothesis: https://t.co/SMDyTiOHTC
 Reply      Retweet   48      Like     228    

More by:   DeepMind
zoom in  
 



DeepMind    @DeepMind   ·   10/5/2021
Many single protein chains can be accurately predicted using AlphaFold, but multi-chain protein complexes remain challenging. Introducing AlphaFold-Multimer, a model that can predict the structure of multi-chain protein complexes with increased accuracy https://t.co/KMyN87wgDE 1/
 
 Reply      Retweet   50      Like     211    

More by:   DeepMind
zoom in  
 



DeepMind    @DeepMind   ·   7/19/2021
At #ICML 2011, researchers presented a paper that described an early method for Bayesian learning from large scale data that combined stochastic gradient descent, which is used in deep learning, & Langevin dynamics, a basic Monte Carlo method for simulating from a distribution 1/
 Reply      Retweet   26      Like     166    

More by:   DeepMind
zoom in  
 



DeepMind    @DeepMind   ·   9/29/2021
From packing an umbrella to preparing for extreme conditions, predicting short term weather patterns is crucial for daily life. New research with the @metoffice and SOTA model advances the science of Precipitation Nowcasting - the prediction of rain: https://t.co/gKGx76zwxS 1/4
 Reply      Retweet   29      Like     128    

More by:   DeepMind
zoom in  
 



DeepMind    @DeepMind   ·   7/6/2021
Like Transformers, Perceivers process inputs using attention. But unlike Transformers, they first map inputs to a small latent space where processing is cheap & doesn’t depend on the input size. This allows us to build deep networks even when using large inputs like images. (2/)
 Reply      Retweet   13      Like     80    

More by:   DeepMind
zoom in  
 



DeepMind    @DeepMind   ·   9/10/2021
In the first lecture of the series, Research Scientist Hado introduces the course and explores the fascinating connection between reinforcement learning and artificial intelligence: https://t.co/WvlkYjcxTh #DeepMindxUCL @ai_ucl
 Reply      Retweet   16      Like     66    

More by:   DeepMind
zoom in  
 



DeepMind    @DeepMind   ·   10/5/2021
Biological processes are often underpinned by the formation of protein complexes. We hope this work will act as a stepping stone towards executing on more complex folds, such as RNA & DNA molecules. Work by @richevans_dm, Michael O’Neill, @AlexPritzel, @NatashaAnt1, et al. 2/2
 Reply      Retweet   4      Like     62    

More by:   DeepMind
zoom in  
 



DeepMind    @DeepMind   ·   10/4/2021
Enformer aims to better understand variants in the non-coding genome and can predict the effects of variants on gene expression in both natural genetic and synthetic variants. Explore the model and its initial predictions of common genetic variants: https://t.co/fTejl9GYGx 2/
 Reply      Retweet   7      Like     43    

More by:   DeepMind
zoom in  
 



DeepMind    @DeepMind   ·   6/17/2021
Our deep GNNs leverage recent DeepMind research such as BGRL (BYOL for graphs) & Noisy Nodes (denoising GNN regulariser). The hope is that this work can have an immediate impact on large-scale applications of GNNs, especially for social networks & computational chemistry. (2/)
 Reply      Retweet   4      Like     56    

More by:   DeepMind
zoom in  
 



DeepMind    @DeepMind   ·   10/11/2021
Introducing RGB-Stacking, a new benchmark where the goal is to train a robotic arm - using reinforcement learning - to stack real objects of diverse shapes that elicit complex object dynamics: https://t.co/q4nHaIu3rr Paper https://t.co/wXX8g8EEEn Video https://t.co/2ezY1Nyrk5 1/
 Reply      Retweet   6      Like     14    

More by:   DeepMind
zoom in  
 



DeepMind    @DeepMind   ·   9/29/2021
Today’s weather systems provide planet-scale predictions several days ahead, but often struggle to generate high-resolution predictions for short lead times. Nowcasting fills this performance gap, with predictions on rainfall within the next 1-2 hours. 2/4
 Reply      Retweet   1      Like     21    

More by:   DeepMind
zoom in  
 



DeepMind    @DeepMind   ·   10/12/2021
Does Bayesian deep learning work? The Neural Testbed provides tools to evaluate uncertainty estimates. These tools assess both the quality of marginal prediction per input & joint predictions given many inputs. Github: https://t.co/0xdiYiiDwQ Paper: https://t.co/UFgXYwf6Ln 1/
 
 Reply      Retweet   1      Like     19    

More by:   DeepMind
zoom in  
 









  More from:  


Quanta Magazine
Big ideas in science and math. Because you want to know more. Launched by @SimonsFdn. https://t.co/n6itBXqN3A Get our weekly newsletter: https://t.co/XI6i48Zcdk

Yann LeCun
Professor at NYU. Chief AI Scientist at Facebook. Researcher in AI, Machine Learning, etc. ACM Turing Award Laureate.

Google AI
Google AI is focused on bringing the benefits of AI to everyone. In conducting and applying our research, we advance the state-of-the-art in many domains.

Facebook AI
Facebook AI focuses on bringing the world together by advancing AI, powering meaningful and safe experiences, and conducting open research.

Demis Hassabis
Founder & CEO @DeepMind - #AlphaFold, #AlphaGo, #AlphaZero & Atari DQN. Working on General AI. Trying to understand what is *really* going on in the universe

Sebastian Raschka
Author of the 'Python Machine Learning' book. Tweet about Python, deep learning research, open source. Asst Prof of Statistics @UWMadison. Opinions are my own.