Sokoban

Description

Sokoban is an environment for evaluating agents on spatial planning and sequential puzzle solving. This environment wraps the Sokoban implementation from TextArena, a framework for text-based game environments.

Capabilities

• Long-term spatial planning and reasoning
• Understanding irreversible actions and dead states
• Multi-step goal decomposition
• Strategic path planning with obstacles

Compute Requirements

Sokoban does not require a sandbox. It has minimal compute requirements.

License

MIT.

Tasks

There are two splits: train (300 tasks) and test (300 tasks). Each split contains 50 tasks across each of 6 variants:

• Sokoban-v0
• Sokoban-v0-medium
• Sokoban-v0-medium-raw
• Sokoban-v0-medium-train
• Sokoban-v0-raw
• Sokoban-v0-train

Each task is seeded for reproducibility.

Reward Structure

This is a sparse reward environment. Rewards are mapped from TextArena's native range of {-1, 0, 1} to {0.0, 0.5, 1.0} via (raw + 1) / 2.

We do not use LLM graders for this environment; reward is determined programmatically.

Data

Game state is generated procedurally by the TextArena engine using seeded randomness. No external data files are required.

Tools

Agents are given a single tool:

• move(direction): Move in the given direction (up, down, left, right). Push boxes by moving into them.

Time Horizon

Sokoban is a multi-turn environment.

Environment Difficulty

This environment ranges from moderate to challenging depending on the variant. The medium variant presents increased difficulty with more complex level layouts and additional boxes to manage.

Other Environment Requirements

There are no further environment requirements; Sokoban works out of the box without any secrets or API keys.

Safety

Agents in Sokoban interact only with a puzzle game and have no access to external systems, the internet, or sensitive data. The environment does not present safety risks.

Citations

@software{textarena2024,
  author    = {Guertler, Leon and Banting, Wilfried and Pignatelli, Eduardo},
  title     = {TextArena},
  year      = {2024},
  publisher = {GitHub},
  url       = {https://github.com/LeonGuertler/TextArena}
}