S

Suspicion-Agent

AI agent for playing imperfect information games with GPT-4.

AI AgentOpen SourceEarly
Visit Website160

What is Suspicion-Agent?

Suspicion-Agent is aI agent for playing imperfect information games with GPT-4.

About

Suspicion-Agent is designed to engage in imperfect information games using a Theory of Mind approach with GPT-4. It allows developers to train and evaluate AI agents in various game scenarios, making it suitable for researchers and developers interested in game theory and AI interactions. Key capabilities include agent training, evaluation against baselines, and visualizing game data.

Strengths

  • Utilizes advanced GPT-4 capabilities for game interactions.
  • Supports multiple training models (CFR, NFSP, DQN, DMC).
  • Provides visualizations for game data and agent performance.

Limitations

  • Requires OpenAI API key for functionality.
  • Dependent on Python environment setup.
  • Limited to specific game types and scenarios.

Use Cases

Training AI agents for strategic decision-making in games.Evaluating agent performance against baseline models.Visualizing game data for analysis and improvement.

Integrations

OpenAI APIHugging Face SpacesSkyAGI