Neural Networks for Chess

TL;DR

This book provides a comprehensive introduction to neural networks and search techniques used in modern chess engines like AlphaZero, covering their technical foundations, design, and implementation through practical examples.

Contribution

It offers a detailed, educational overview of neural network architectures and classical search algorithms specific to computer chess, including a miniaturized AlphaZero implementation for learning purposes.

Findings

Neural networks form the core of modern chess engines.

AlphaZero's self-play reinforcement learning outperforms supervised training.

Miniature AlphaZero can be implemented using Hexapawn as an example.

Abstract

AlphaZero, Leela Chess Zero and Stockfish NNUE revolutionized Computer Chess. This book gives a complete introduction into the technical inner workings of such engines. The book is split into four main chapters -- excluding chapter 1 (introduction) and chapter 6 (conclusion): Chapter 2 introduces neural networks and covers all the basic building blocks that are used to build deep networks such as those used by AlphaZero. Contents include the perceptron, back-propagation and gradient descent, classification, regression, multilayer perceptron, vectorization techniques, convolutional networks, squeeze and excitation networks, fully connected networks, batch normalization and rectified linear units, residual layers, overfitting and underfitting. Chapter 3 introduces classical search techniques used for chess engines as well as those used by AlphaZero. Contents include minimax, alpha-beta search, and Monte Carlo tree search. Chapter 4 shows how modern chess engines are designed. Aside from the ground-breaking AlphaGo, AlphaGo Zero and AlphaZero we cover Leela Chess Zero, Fat Fritz, Fat Fritz 2 and Efficiently Updatable Neural Networks (NNUE) as well as Maia. Chapter 5 is about implementing a miniaturized AlphaZero. Hexapawn, a minimalistic version of chess, is used as an example for that. Hexapawn is solved by minimax search and training positions for supervised learning are generated. Then as a comparison, an AlphaZero-like training loop is implemented where training is done via self-play combined with reinforcement learning. Finally, AlphaZero-like training and supervised training are compared.