Empirical evaluation of AlphaGo Zero. a Performance of self-play

Mastering the game of Go without human knowledge

A (Long) Peek into Reinforcement Learning

RLiable: Towards Reliable Evaluation & Reporting in Reinforcement Learning – Google Research Blog

AI versus AI: Self-Taught AlphaGo Zero Vanquishes Its Predecessor

Empirical evaluation of AlphaGo Zero. a Performance of self-play

A (Long) Peek into Reinforcement Learning

Student of Games: A unified learning algorithm for both perfect and imperfect information games

A general reinforcement learning algorithm that masters chess, shogi, and Go through self-play

neural network - AlphaGo Zero board evaluation function uses multiple time steps as an input Why? - Stack Overflow