Show HN: NarayanaDB – 认知数据库与通用人工智能研究平台
1 分•作者: railspress•7 个月前
嗨，HN！
我构建了 NarayanaDB，一个高性能的列式数据库，同时也是一个认知型通用人工智能（AGI）平台。它集成了：
* 意识持续循环（CPL）——用于推理的持续思考循环
* 记忆系统——情景记忆、语义记忆、程序性记忆和工作记忆
* 叙事身份——特质、遗传和个性建模
* 道德推理——可选的“会说话的蟋蟀”层，用于伦理决策
* 世界接口——智能体的感官和运动交互
* LLM 集成——推理、记忆总结、原则演化
它完全模块化、可插拔，专为希望试验具有道德意识的学习型智能体的研究人员、AGI 爱好者和开发人员设计。
代码库和文档：[https://github.com/carlosbarbosamexico/narayana](https://github.com/carlosbarbosamexico/narayana)
很想听听你们的想法、反馈或潜在的应用案例。
查看原文
Hi HN!<p>I built NarayanaDB, a high-performance columnar database that doubles as a cognitive AGI platform. It integrates:
• Conscience Persistent Loop (CPL) – continuous thought loop for reasoning
• Memory systems – episodic, semantic, procedural, working memory
• Narrative identity – traits, genetics, and personality modeling
• Moral reasoning – optional Talking Cricket layer for ethical decision-making
• World interface – sensory and motor interaction for agents
• LLM integration – reasoning, memory summarization, principle evolution<p>It’s fully modular, pluggable, and designed for researchers, AGI enthusiasts, and developers who want to experiment with morally-aware, learning agents.<p>Repo & docs: <a href="https://github.com/carlosbarbosamexico/narayana" rel="nofollow">https://github.com/carlosbarbosamexico/narayana</a><p>Curious to hear your thoughts, feedback, or potential use cases.