Large language models, a cornerstone of modern AI, are reshaping industries and accelerating innovation across diverse domains. At Sea AI Lab, we conduct research to develop smarter language models and transform them into productivity. We focus on three complementary research directions:

Advancing base models: We conduct fundamental research to develop better language models with enhanced understanding, reasoning, and knowledge generation abilities.

Enhancement via post training: We develop post-training techniques that allow models to further improve, for the purpose of enhancing their both general and specific problem-solving capabilities beyond initial training.

Solving real-world challenges: We build systems that leverage our research to address complex problems, creating solutions that help people work smarter and accomplish more.