Learning the Latent Rules of a Game from Data: A Chess Story

TL;DR

This paper shows that small pretrained language models can learn the rules of chess, suggest legal moves, and solve problems effectively through instruction fine-tuning with a relatively small dataset.

Contribution

It demonstrates that small foundational language models can learn complex game rules and strategies from data, with improved accuracy through fine-tuning and increased examples.

Findings

Small models (28M and 125M parameters) can learn chess rules from data.

Fine-tuning with more examples reduces model hallucinations.

Successive fine-tuning epochs improve model performance.

Abstract

We demonstrate that small pretrained foundational generative language models with millions of parameters can learn the latent rules of a process from data associated with the process. Inspired by Stefan Zweig's novella "Schachnovelle," also known as "The Royal Game" in English, we show that 28M and 125M parameter pretrained foundational small language models (SLMs) can be instruction fine-tuned with 1,000-to-1,000,000 examples to learn the rules of chess, propose legal moves, and accurately solve chess problems. We also explore the impact of successive language model fine-tuning epochs on improved outcomes and demonstrate reductions in model hallucinations by increasing the number of instruction fine-tuning examples.