Puzzle Solving without Search or Human Knowledge: An Unnatural Language Approach

Understand

The application of Generative Pre-trained Transformer (GPT-2) to learn text-archived game notation provides a model environment for exploring sparse reward gameplay.

• The transformer architecture proves amenable to training on solved text archives describing mazes, Rubik's Cube, and Sudoku solvers.
• The method benefits from fine-tuning the transformer architecture to visualize plausible strategies derived outside any guidance from human heuristics or domain expertise.
• The large search space ($>10^{19}$) for the games provides a puzzle environment in which the solution has few intermediate rewards and a final move that solves the challenge.