Can better architectures & representations make self-play enough for zero-shot coordination? ๐ค
We explore this in our ICLR 2025 paper: A Generalist Hanabi Agent. We develop R3D2, the first agent to master all Hanabi settings and generalize to novel partners! ๐ #ICLR2025 1/n
04.04.2025 17:12
๐ 13
๐ 4
๐ฌ 1
๐ 3