Shanghai has received national recognition as one of the top 10 typical cases in the third batch of data intellectual property registrations. The Shanghai Intellectual Property Administration recommended the "Kupas High-Quality Education Thinking Chain Dataset" case.
Developed jointly by Shanghai Kupas Technology Co Ltd, Shanghai Jiao Tong University, Fudan University, and 12 other universities and institutions, the dataset covers higher mathematics, physics, and computer science. It follows a structured workflow from problem standardization to reasoning chain generation, annotation, quality checks, and version management.
The dataset features an innovative iterative reflection prompt system and type-specific quality checks. It now includes 517,000 mathematical and engineering reasoning chains and a general corpus of 178.49 TB. Open-source downloads have exceeded 1.5 million, serving nearly 500,000 users and offering a replicable "Education+AI" solution.
Shanghai, as one of China's first pilot areas for data intellectual property, has seen steady progress since the implementation of the Interim Measures for the Registration and Evidence Preservation of Data Product Intellectual Property in December 2024. Market entities have filed more than 1,200 registrations.
Looking ahead, the Shanghai Intellectual Property Administration will continue to deepen the pilot program, cultivate more high-quality cases, and support the high-quality development of Shanghai's digital economy.
