IteratedPrisonersDilemma

Description

IteratedPrisonersDilemma is an environment for evaluating agents on cooperation and defection strategies in the classic game theory dilemma. This environment wraps the IteratedPrisonersDilemma implementation from TextArena, a framework for text-based game environments.

Capabilities

• Game theory and cooperation strategies
• Reputation building and trust dynamics
• Strategic adaptation over multiple rounds
• Competitive gameplay against an LLM opponent

Compute Requirements

IteratedPrisonersDilemma 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:

• IteratedPrisonersDilemma-v0
• IteratedPrisonersDilemma-v0-train
• IteratedPrisonersDilemma-v0-raw

Each task is seeded for reproducibility.

Reward Structure

This is a sparse reward environment. Rewards are mapped from TextArena's native range to [0.0, 1.0] via max(0.0, min(1.0, (raw + 50) / 100)).

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 two tools:

• cooperate(params): Choose to cooperate this round.
• defect(params): Choose to defect this round.

Time Horizon

IteratedPrisonersDilemma is a multi-turn environment.

Environment Difficulty

Medium - requires balancing short-term gains from defection with long-term benefits of cooperation, while adapting to opponent behavior patterns.

Other Environment Requirements

This environment requires an OpenAI API key (passed via secrets) to power the LLM opponent.

Safety

Agents in IteratedPrisonersDilemma interact only with a game theory simulation 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}
}