x-glacier/GenerativeAgentsCN
A simulation of 25 AI agents living autonomously in a virtual town, powered by LLMs like Qwen3 and DeepSeek-R1.
Not currently ranked — collecting fresh signals.
star history
This project reconstructs and localizes Stanford’s Generative Agents into a maintainable Chinese version. Twenty-five AI agents, driven entirely by large language models, autonomously organize social activities including parties and meetings within a simulated virtual environment. It supports local deployment via Ollama, uses Pydantic models for structured output parsing, and integrates Chinese-optimized prompts for models such as Qwen and GLM-4.
Frequently asked
- What is x-glacier/GenerativeAgentsCN?
- A simulation of 25 AI agents living autonomously in a virtual town, powered by LLMs like Qwen3 and DeepSeek-R1.
- Is GenerativeAgentsCN open source?
- Yes — x-glacier/GenerativeAgentsCN is open source, released under the Apache-2.0 license.
- What language is GenerativeAgentsCN written in?
- x-glacier/GenerativeAgentsCN is primarily written in Python.
- How popular is GenerativeAgentsCN?
- x-glacier/GenerativeAgentsCN has 500 stars on GitHub.
- Where can I find GenerativeAgentsCN?
- x-glacier/GenerativeAgentsCN is on GitHub at https://github.com/x-glacier/GenerativeAgentsCN.