Underlying paper - Generative Agents

Overview

This paper introduces generative agents, computational software agents that simulate believable human behavior through an architecture supporting daily routines, memory, reflection, and social interactions. Agents autonomously wake up, cook breakfast, work, form opinions, and initiate conversations, remembering past events to plan future actions.

Best for

Best for
Researchers and developers building human-like social simulations in games, VR, or prototyping tools

Use cases

• Building realistic NPCs for immersive games and virtual worlds
• Prototyping social simulations for interpersonal communication research
• Creating generative personas for role-playing or rehearsal environments

Notes

This paper introduces generative agents, computational software agents that simulate believable human behavior through an architecture supporting daily routines, memory, reflection, and social interactions. Agents autonomously wake up, cook breakfast, work, form opinions, and initiate conversations, remembering past events to plan future actions.

Use cases

• Building realistic NPCs for immersive games and virtual worlds
• Prototyping social simulations for interpersonal communication research
• Creating generative personas for role-playing or rehearsal environments

Pros

• Provides a concrete, published architecture for long-term memory and planning
• Enables emergent social behaviors through agent-to-agent interactions
• Openly available as a research paper with detailed implementation insights

Cons

• A research paper, not a ready-to-use library or SDK
• Requires significant implementation effort to reproduce the system
• Computational cost may be high when simulating many agents concurrently

Indexed from awesome-ai-agents and enriched against its public facts.