Understanding AlphaZero Neural Network's SuperHuman Chess Ability

As a common and (sometimes) proven belief, deep learning systems seem to learn uninterpretable representations and are far from human understanding. Recently, some studies have highlighted the fact that this may not always be applicable, and some networks may be able to learn human-readable representations. Unfortunately, this ability could merely come from the fact that these networks are exposed to human-generated data. So, to demonstrate their ability to learn like humans (and not that they are simply memorizing human-created labels), it is necessary to test them without any label.  Following this idea, the DeepMind and Google Brain teams, together with

What's Inside AlphaZero's Chess Brain?

Alpha Zero's “Alien” Chess Shows the Power, and the Peculiarity, of AI

Chess AI: A Brief History

Even Superhuman Go AIs Have Surprising Failures Modes – Center for Human-Compatible Artificial Intelligence

AlphaDDA: strategies for adjusting the playing strength of a fully trained AlphaZero system to a suitable human training partner [PeerJ]

How AlphaZero Learns Chess

Electronics, Free Full-Text

Google's AlphaZero AI mastered chess in just four hours

[Sadler, Matthew, Regan, Natasha, Kasparov, Garry] on . *FREE* shipping on qualifying offers. Game Changer: AlphaZero's Groundbreaking

Game Changer: AlphaZero's Groundbreaking Chess Strategies and the Promise of AI

A Beginner's Guide to Machine Learning - RapidMiner

Google AlphaZero Beats The Best Chess-Playing Software Program