Game-playing by Computers

Papers

• 2017 Silver et all, Mastering the game of Go without human knowledge (AlphaGo Zero)
• 2017 Silver et all, Mastering Chess and Shogi by Self-Play with a General Reinforcement Learning Algorithm (Alpha Zero)
• 2019 David Wu (Katago), Accelerating Self-Play Learning in Go
• 2020 Silver et all, Mastering Atari, Go, Chess and Shogi by Planning with a Learned Model (Mu Zero)
• 2021 Weirui Ya et al, Mastering Atari Games with Limited Data (Efficient Zero; see also github and lesswrong)
• 2022 Silver et al, Planning in Stochastic Environments with a Learned Model (Stochastic MuZero, github)

Interesting Repositories

• KataGo
• Alpha Zero General and a blog post
• AlphaZero.jl (Julia) and its excellent docs
• MuZero General (might be useful for like Wesnoth?)
• OpenSpiel … not really sure what that is, but they sure implemented all the games

Less popular but still interesting:

• alpha-zero-boosted (w/ Gradient Boosted Decision Trees) implements TwixT! some discussion here…