Skip to main content

Agent Simulations

Agent simulations involve taking multiple agents and having them interact with each other.

They tend to use a simulation environment with an LLM as their "core" and helper classes to prompt them to ingest certain inputs such as prebuilt "observations", and react to new stimuli.

They also benefit from long-term memory so that they can preserve state between interactions.

Like Autonomous Agents, Agent Simulations are still experimental and based on papers such as this one.

Help us out by providing feedback on this documentation page: