In a collaboration with Google Brain, we are releasing the Hanabi Learning Environment: a testbed for collaborative multi-agent learning and theory of mind in the card game Hanabi. Paper: GitHub:
Google AI and researchers have open-sourced the Hanabi Learning Environment, a testbed for collaborative multi-agent learning research. Check out the paper at and the code at
Environment and agent code using Dopamine are available on GitHub, so you can get started on these challenges right away!