AlphaStar: Grandmaster level in StarCraft II using multi-agent reinforcement learning

[Rémi Coulom]

https://www.deepmind.com/blog/article/A ... t-learning

Direct link to the paper: https://storage.googleapis.com/deepmind ... matted.pdf

The most interesting contribution seems to be their league system to reach the Nash equilibrium. Their method for off-policy learning is very interesting, too. This gives me some inspiration for my mahjong reinforcement-learning experiments.