SAI, a Sensible Artificial Intelligence that plays Go
This is an incremental improvement for Go AI enthusiasts and researchers, as it enhances reinforcement learning efficiency in a specific domain.
The authors tackled the problem of improving Go-playing AI by modifying the AlphaGo Zero paradigm to handle multiple komi values, resulting in a very strong playing agent on 7x7 Go.
We propose a multiple-komi modification of the AlphaGo Zero/Leela Zero paradigm. The winrate as a function of the komi is modeled with a two-parameters sigmoid function, so that the neural network must predict just one more variable to assess the winrate for all komi values. A second novel feature is that training is based on self-play games that occasionally branch -- with changed komi -- when the position is uneven. With this setting, reinforcement learning is showed to work on 7x7 Go, obtaining very strong playing agents. As a useful byproduct, the sigmoid parameters given by the network allow to estimate the score difference on the board, and to evaluate how much the game is decided.