cesure

[ PAPER ] · 2019 · Nature

Grandmaster level in StarCraft II using multi-agent reinforcement learning

O. Vinyals, Igor Babuschkin, Wojciech M. Czarnecki, Michaël Mathieu, A. Dudzik, Junyoung Chung, David Choi, Richard Powell, T. Ewalds, Petko Georgiev, Junhyuk Oh, Dan Horgan, M. Kroiss, Ivo Danihelka, Aja Huang, L. Sifre, Trevor Cai, J. Agapiou, Max Jaderberg, A. Vezhnevets, R. Leblond, Tobias Pohlen, Valentin Dalibard, D. Budden, Yury Sulsky, James Molloy, T. Paine, Caglar Gulcehre, Ziyun Wang, T. Pfaff, Yuhuai Wu, Roman Ring, Dani Yogatama, Dario Wünsch, Katrina McKinney, Oliver Smith, T. Schaul, T. Lillicrap, K. Kavukcuoglu, D. Hassabis, C. Apps, David Silver

4,417 citationsno open PDF · spoken briefing availableDOI
No open-access PDF, so we'll narrate a briefing from the abstract instead.

[ TLDR ]

The agent, AlphaStar, is evaluated, which uses a multi-agent reinforcement learning algorithm and has reached Grandmaster level, ranking among the top 0.2% of human players for the real-time strategy game StarCraft II.