Collected sources and patterns will appear here. Add from search, explore, or the patterns library.
GameState -> Tensor
Convert a player's private and public observation history in an imperfect-information game state into a normalized numeric tensor.
Problem it solves
Neural network-based reinforcement learning agents require fixed-size numeric representations of partially observable states.
Consumes
Emits
The real projects this mechanism was found in. Attribution is the point — this is how the best teams actually do it.