Exploring multiple paths toward Artificial General Intelligence.
AGI-Brain is an open research and engineering group dedicated to the pursuit of Artificial General Intelligence (AGI) 🚀
We are interested in any approach, methodology, or system that contributes to the understanding and construction of general intelligent agents.
Our research spans a broad range of directions, including but not limited to:
- 🤖 Embodied intelligence and robotics
- 🎯 Decision-making and control
- 🧠 Reinforcement learning and sequential reasoning
- 👁️ Machine perception and representation learning
- 🌍 Learning, planning, and interaction in complex environments
Rather than committing to a single paradigm, AGI-Brain embraces multiple complementary research paths, combining theoretical insights, empirical exploration, and scalable engineering systems.
Our long-term goal is to build reusable, extensible, and reproducible infrastructures 🧩 that support sustained research toward general-purpose intelligence.
- XuanCe – A unified and extensible framework for reinforcement learning and agent-based research
- XuanCe-Benchmarks – Standardized evaluation and benchmarking for agent learning systems