STAIR
We are forcusing on scalable and trustworthy AI researches, which includes trustworthy foundation models and its ability scaling, and scalable algorithms especially on big graphs, aiming at bridging large language models (LLMs) and big graph mining to tackle the real-world challenges.
The group is directed by Shenghua Liu at Institute of Computing Technology, Chinese Academy of Sciences (ICT, CAS).
Our goals include:
Trustworthy LLMs, focusing on knowledge editing, contextual faithfulness, and safe generation.High-quality data, that is data mining methods related to big graph mining, scalable algorithms, graph compression, and summarization.Reasoning and efficiency scaling, focusing on more complex reasoning abilities (like thinking as a graph, and graph LLMs), and efficient machine learning from both algorithm and hardware sides.
We are actively recruiting highly-motivated Researchers and Ph.D./M.S./visiting students to join our group at ICT, CAS!