Intelligence
Exploring how learning systems can become more capable, useful, and adaptable across different tasks.
OpenFSAI / Research
Research for future intelligence.
We study the foundations of increasingly capable AI systems, with a focus on useful behavior, reliable evaluation, and responsible development.
Evaluation
Building clearer ways to understand what systems can do, where they fall short, and how progress should be measured.
Safety
Keeping responsibility, transparency, and human benefit central as AI capabilities advance.