Artificial Intelligence

Beyond models and predictions

Yes, large-scale language models and generative artificial intelligence are hot, but we believe AI is about more than models and predictions. Our research focuses on methodology, which is about thinking itself. We strive to build AI that can be trusted and conduct research responsibly.

We are committed to research that benefits all humanity

If machines can independently do scientific research based on human knowledge and produce new knowledge or valuable results, it will be a new era of human civilization.

We empower robots with real-world skills

We teach robots problem-solving methodologies and solve real-world problems step by step.

Security is the bottom line

We conduct responsible research to build AI you can trust. We integrate risk control into our workflow from the very beginning of our design.

Join us in shaping the future of technology.