Two images showing RETRO Architecture and Gopher (280B) vs State of the Art
DeepMind

Large Language Models Shrink: Gopher and RETRO prove lean language models can push boundaries.

DeepMind released three papers that push the boundaries — and examine the issues — of large language models.
Animated illustration shows the model architecture of a graph neural network.
DeepMind

A Deeper Look at Graphs: Graph Neural Networks Work Better With More Layers

New research shows that drastically increasing the number of layers in a graph neural networks improves its performance on large datasets.
An animation of a protein modeled by AlphaFold 2.
DeepMind

DeepMind Doubles Down on AlphaFold: DeepMind Launches Company to Commercialize AlphaFold 2

Isomorphic aims to build its business on AlphaFold 2, an ensemble of neural networks that finds the shapes of protein molecules.
Animation showing a simulated football team and how it works
DeepMind

Team Players: Football-Playing AI Blends Individual and Group Skills

Playing a team sport involves a fluid blend of individual and group skills. Researchers integrated both types of action into realistic humanoid agents that play football (known as soccer in the U.S.).
Results of survey about how AI Engineers vs US public feel about ethical issues
DeepMind

AI Engineers Weigh In on AI Ethics: Survey Shows How AI Engineers Feel About Ethical Issues

Machine learning researchers tend to trust international organizations, distrust military forces, and disagree on how much disclosure is necessary when describing new models, a new study found.
Sequence of famous arcade games' scenes
DeepMind

Solve RL With This One Weird Trick: How to get better performance from reinforcement learning.

The previous state-of-the-art model for playing vintage Atari games took advantage of a number of advances in reinforcement learning (RL). The new champion is a basic RL architecture plus a trick borrowed from image generation.
Animation showing AlphaFold working
DeepMind

Biomedical Treasure Chest

DeepMind opened access to AlphaFold, a model that finds the shapes of proteins, and to its output so far — a potential cornucopia for biomedical research. The research lab, a division of Google’s parent company Alphabet, made AlphaFold freely available.
Series of images showing how single trained network generates 3D reconstructions of multiple scenes
DeepMind

One Network, Many Scenes

To reconstruct the 3D world behind a set of 2D images, machine learning systems usually require a dedicated neural network for each scene. New research enables a single trained network to generate 3D reconstructions of multiple scenes.
Animation of SourceAI working
DeepMind

Robocoders

Language models are starting to take on programming work. SourceAI uses GPT-3 to translate plain-English requests into computer code in 40 programming languages. The French startup is one of several companies that use AI to ease coding.
Graphs and data related to recurrent neural nets (RNNs)
DeepMind

Performance Guaranteed

Bayes-optimal algorithms always make the best decisions given their training and input, if certain assumptions hold true. New work shows that some neural networks can approach this kind of performance.
Different chess moves
DeepMind

Chess: The Next Move

AI has humbled human chess masters. Now it’s helping them take the game to the next level. DeepMind and retired chess champion Vladimir Kramnik trained AlphaZero, a reinforcement learning model that bested human experts in chess, Go, and Shogi, to play-test changes in the rules.
Data related to experience replay
DeepMind

Experience Counts

If the world changes every second and you take a picture every 10 seconds, you won’t have enough pictures to observe the changes clearly, and storing a series of pictures won’t help. On the other hand, if you take a picture every tenth of a second, then storing a history will help model the world.
Data related to AlphaFold, a protein-folding model
DeepMind

Protein Shapes Revealed

A protein’s biological function depends largely on its three-dimensional shape, but deducing its shape from its sequence of amino acids has been a longstanding problem. Researchers at DeepMind reveal how they used deep learning to solve the puzzle.
Maze action video game Pac-Man
DeepMind

Two-Way Winner

AlphaGo Zero demonstrates superhuman performance playing Go, chess, and shogi. Models like R2D2 do the same playing classic Atari titles. A new approach to deep reinforcement learning is the first to achieve state-of-the-art results playing both board and video games.
Breast cancer screening
DeepMind

Cancer in the Crosshairs

Computer vision has potential to spot cancer earlier and more accurately than human experts. A new system surpassed human accuracy in trials, but critics aren’t convinced.

Subscribe to The Batch

Stay updated with weekly AI News and Insights delivered to your inbox