Search
On September 13, Zhipu made headlines by announcing the successful closure of a funding round amounting to nearly $5 billion. The substantial capital injection is earmarked for the advancement of its next-generation GLM foundation model, the development of a fully autonomous training system, and the construction of essential computational infrastructure. Zhipu describes its fully autonomous training approach as a method where the next-generation GLM model is trained within the framework established by its predecessor, fostering a cycle of recursive self-improvement. Specific areas of investment encompass the automated generation and curation of training data, the creation of specialized task environments, the bolstering of long-range reasoning capabilities, as well as the optimization of domestic chip compatibility, operator development, and inference processes.
-
C114 Communication Network -
Communication Home
