Our team is the first few in this line of research, and have produced a series of works, including (a) bridging core vision tasks with language: UniCL, RegionCLIP, GLIP, KLITE, foundation model Florence; (b) developing generalist decoder X-Decoder, and (c) enabling more promptable and interactive systems like SEEM, Semantic-SAM, LLaVa, and SoM for GPT-4V.