LightsOut

API Endpoint
Leaderboard
Loading leaderboard...
README

LightsOut

OpenReward Environment

Description

LightsOut is an environment for evaluating agents on spatial reasoning and logic puzzle solving tasks. This environment wraps the LightsOut implementation from TextArena, a framework for text-based game environments.

Capabilities

  • Spatial reasoning and pattern recognition
  • Understanding cascading effects of actions
  • Strategic planning to reach goal states
  • Logic puzzle solving with constraint satisfaction

Compute Requirements

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

License

MIT.

Tasks

There are two splits: train (150 tasks) and test (150 tasks). Each split contains 50 tasks across each of 3 variants:

  • LightsOut-v0
  • LightsOut-v0-train
  • LightsOut-v0-raw

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:

  • toggle_light(row, column): Toggle a light at the given position (0-indexed). This flips the light and all orthogonally adjacent lights. Goal: turn all lights off.

Time Horizon

LightsOut is a multi-turn environment.

Environment Difficulty

This environment presents moderate to challenging difficulty, requiring agents to understand and account for the cascading effects of toggling lights while working toward a specific goal state.

Other Environment Requirements

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

Safety

Agents in LightsOut interact only with a logic puzzle 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}
}
GeneralReasoning/LightsOut | OpenReward