WalkingLabs is an open-source lab focused on agentic AI, intelligent agents, and reinforcement learning (RL), from THU SIGS.
- Our goal is to permanently reduce the difficulty for everyone to learn large language models and agents within this year.
If you are interested in agents, RL, LLM systems, embodied intelligence, or next-generation AI infrastructure, welcome to explore and build with WalkingLabs.
中文介绍
WalkingLabs 是一个专注于 Agentic AI、智能体系统(Agents)与强化学习(RL) 的开源实验室。
我们致力于探索能够感知、规划、推理、行动与持续优化的新一代智能体系统,并围绕大模型驱动的 Agent、强化学习、工具使用、多智能体协作以及真实任务场景,打造高质量的开源项目、教程与开发基础设施。
- Agentic AI:面向复杂任务的自主智能体架构与执行系统
- LLM Agents:基于大语言模型的规划、记忆、工具调用与任务分解
- Reinforcement Learning:用于决策优化、策略学习与智能体能力提升的 RL 方法
- Multi-Agent Systems:多智能体协作、通信与任务编排机制
- Open Tutorials & Open Source:以清晰的教程降低门槛,以开源项目推动社区共建
我们希望打造一个开放、严谨且长期主义的技术空间,让前沿的智能体理念不仅易于理解,也能够沉淀为可复现、可扩展、可落地的开源系统。
如果你关注 Agent、RL、LLM Systems、Embodied Intelligence 或下一代 AI 开发基础设施,欢迎关注并参与 WalkingLabs。
