AlphaZero: Shedding new light on chess, shogi, and Go
Assessing Game Balance with AlphaZero: Exploring Alternative Rule Sets in Chess
MuZero: Mastering Go, chess, shogi and Atari without rules
A general reinforcement learning algorithm that masters chess, shogi, and Go through self-play
Mastering Atari, Go, Chess and Shogi by Planning with a Learned Model
Mastering Atari, Go, Chess and Shogi by Planning with a Learned Model
The Podcast: Episode 2: Go to Zero
AlphaZero Resources
The Podcast: Episode 8: Demis Hassabis - The interview
Mastering Stratego, the classic game of imperfect information
Policy improvement by planning with Gumbel
Computations Underlying Social Hierarchy Learning: Distinct Neural Mechanisms for Updating and Representing Self-Relevant Information
Reinforcement Learning with Information Theoretic Actuation
Open-ended Learning in Symmetric Zero-sum Games
Approximate exploitability: Learning a best response in large games
Vector Quantized Models for Planning
Equivariant MuZero
The Hanabi Challenge: A New Frontier for AI Research
Sound Search in Imperfect Information Games
Physically Embedded Planning Problems: New Challenges for Reinforcement Learning
AI for the board game Diplomacy
Introduction to Reinforcement Learning with David Silver
MuZero’s first step from research into the real world
DeepMind’s latest research at ICLR 2022
On the Expressivity of Markov Reward
Using Unity to Help Solve Intelligence
From unlikely start-up to major scientific organisation: Entering our tenth year at DeepMind
DeepMind and Blizzard to release StarCraft II as an AI research environment
Strengthening the AI community
Discovering novel algorithms with AlphaTensor
AlphaStar: Grandmaster level in StarCraft II using multi-agent reinforcement learning
Generally capable agents emerge from open-ended play
A new model and dataset for long-range memory
AlphaStar: Mastering the real-time strategy game StarCraft II
Fast reinforcement learning through the composition of behaviours
International evaluation of an AI system for breast cancer screening
Specifying AI safety problems in simple environments