LGCLMar 21, 2024

Emergent World Models and Latent Variable Estimation in Chess-Playing Language Models

arXiv:2403.15498v257 citationsh-index: 7
Originality Incremental advance
AI Analysis

This work addresses the debate on whether language models extract semantics and world models from text, extending prior research into a more complex domain with real data.

The paper investigates whether language models learn internal world models by training a GPT model on real chess games and finds evidence of internal representations of board state and latent variables like player skill, with interventions using these representations improving the model's win rate by up to 2.6 times.

Language models have shown unprecedented capabilities, sparking debate over the source of their performance. Is it merely the outcome of learning syntactic patterns and surface level statistics, or do they extract semantics and a world model from the text? Prior work by Li et al. investigated this by training a GPT model on synthetic, randomly generated Othello games and found that the model learned an internal representation of the board state. We extend this work into the more complex domain of chess, training on real games and investigating our model's internal representations using linear probes and contrastive activations. The model is given no a priori knowledge of the game and is solely trained on next character prediction, yet we find evidence of internal representations of board state. We validate these internal representations by using them to make interventions on the model's activations and edit its internal board state. Unlike Li et al's prior synthetic dataset approach, our analysis finds that the model also learns to estimate latent variables like player skill to better predict the next character. We derive a player skill vector and add it to the model, improving the model's win rate by up to 2.6 times.

Code Implementations1 repo
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes