Tang Jie
Last reviewed
Jun 9, 2026
Sources
22 citations
Review status
Source-backed
Revision
v3 · 1,925 words
Improve this article
Add missing citations, update stale details, or suggest a clearer explanation.
Suggest edit
CtrlK
Last reviewed
Jun 9, 2026
Sources
22 citations
Review status
Source-backed
Revision
v3 · 1,925 words
Add missing citations, update stale details, or suggest a clearer explanation.
Tang Jie (Chinese: 唐杰; born 1977) is a Chinese computer scientist and a professor in the Department of Computer Science and Technology at Tsinghua University. [1][2] He is known for his academic work on social network mining, knowledge graphs, and data mining, including the AMiner academic search and mining system, and for his role as a co-founder and chief scientist of Zhipu AI, the Beijing startup that commercializes the GLM and ChatGLM family of large language models. [1][3][4] The GLM effort grew out of Tsinghua's Knowledge Engineering Group (KEG), where Tang has spent most of his research career. [4][5]
Tang is a Fellow of the Association for Computing Machinery (ACM), the Institute of Electrical and Electronics Engineers (IEEE), and the Association for the Advancement of Artificial Intelligence (AAAI). [1][6] In January 2026 Zhipu AI became the first major Chinese generative AI startup to list publicly, on the Hong Kong Stock Exchange, and Tang was identified in regulatory filings as a controlling shareholder and the company's chief scientist. [3][7][8]
Tang received his PhD in computer science and technology from Tsinghua University in 2006. [2][9] He joined the faculty of the Department of Computer Science and Technology at Tsinghua, where he became a full professor and a member of the Knowledge Engineering Group, a research lab focused on data mining, knowledge management, and the semantic web. [1][2] His stated research interests include artificial general intelligence, data mining, social networks, machine learning, and knowledge graphs. [1]
Tang has held a series of editorial and program roles in the data mining and web research communities. He served as program committee chair for The Web Conference (WWW) in 2018 and for CIKM in 2016, as program committee chair for WSDM in 2015, and as an executive editor of the journal ACM Transactions on Knowledge Discovery from Data. [2] He received the National Science Fund for Distinguished Young Scholars, one of China's principal grants for early career researchers. [1][2] Within China's AI research community he has also served in a leadership capacity at the Beijing Academy of Artificial Intelligence, a state backed research institute, and he was one of the technical leads on its Wudao large model project. [4][10]
Tang is the principal builder of AMiner, an academic search and social network mining system formerly known as ArnetMiner. [9][11] The system was launched in March 2006 and indexes researchers, publications, citations, and collaboration networks, then applies mining algorithms to model academic influence and expertise. [9][11] By the mid 2020s AMiner described an index of more than 130 million researcher profiles and roughly 200 million or more papers, with users drawn from over 220 countries and regions. [1][11] The underlying paper, "ArnetMiner: Extraction and Mining of Academic Social Networks," published at the ACM SIGKDD conference in 2008, received the SIGKDD Test of Time Award for Applied Science in 2020, a recognition given to work whose impact has lasted about ten years. [6][12]
Much of Tang's earlier research concerns extracting structured knowledge from large heterogeneous networks. His ACM Fellow citation credits contributions to information and social network mining, and his IEEE Fellow citation credits contributions to knowledge discovery from data and social network mining. [6] This line of work covers topics such as influence analysis in social networks, expert finding, name disambiguation across large bibliographic datasets, and representation learning on graphs, with the AMiner platform serving as both a product and a testbed for the methods. [1][9]
From around 2020 Tang's group at Tsinghua turned to large scale language model pretraining. The group developed the General Language Model, or GLM, a pretraining approach based on autoregressive blank infilling that is designed to handle both natural language understanding and generation tasks within one framework. [4][13] GLM became the architectural basis for the models that Zhipu AI later released. The group scaled the approach to a 130 billion parameter bilingual model, GLM-130B, described in the paper "GLM-130B: An Open Bilingual Pre-trained Model," which lists Tang among its authors and was accepted at the International Conference on Learning Representations (ICLR) in 2023. [13] The model was trained for English and Chinese and was released as open source weights for research and commercial use. [13][14]
The GLM family expanded into a set of conversational and multimodal systems associated with Tsinghua KEG and Zhipu AI. [4][15] Following the public attention on dialogue systems in late 2022, the group built aligned chat models on top of GLM. ChatGLM-6B, a six billion parameter bilingual dialogue model, was open sourced in 2023 and was downloaded a large number of times, which helped popularize the line among Chinese developers. [1][15] Subsequent releases included larger and more capable systems, summarized in the paper "ChatGLM: A Family of Large Language Models from GLM-130B to GLM-4 All Tools," published in 2024, which traces the lineage from GLM-130B through GLM-4. [15] GLM-4 was released in June 2024, and later iterations such as GLM-4.5 and GLM-5 followed under the Zhipu and Z.ai brands. [15][16] The group also produced related open models for code and for image and video generation, including CodeGeeX, CogView, and CogVideo. [1]
Zhipu AI traces its origin to June 2019, when a company was spun out of Tsinghua's Knowledge Engineering Group and the AMiner project. [3][4] It was co-founded by Tang Jie and Li Juanzi, both professors in Tsinghua's Department of Computer Science and Technology, and the legal entity is registered in Beijing as Zhipu Huazhang, marketed overseas as Z.ai and presented in filings under the name Knowledge Atlas Technology. [3][4][8] In its early period the firm concentrated on knowledge graphs and academic AI before shifting its focus to large language models. [4]
Tang serves as the company's chief scientist, with responsibility for model architecture and long range research direction, while Zhang Peng leads the company as chief executive. [4][8] Reporting and IPO documents have described Tang as a controlling shareholder of the company. [7][8] On the day of the Hong Kong listing, Tang issued an internal letter calling for a renewed focus on foundational model research and signaling the coming release of a new generation model, GLM-5. [16][17] Tang is also widely credited as a mentor to a cohort of Chinese AI founders, among them Yang Zhilin of the rival startup Moonshot AI. [4]
Zhipu AI is frequently grouped with a small set of well funded Chinese large model startups, sometimes called the AI tigers or, in Chinese coverage, the six little dragons. [3][18] The company had raised on the order of 1.5 billion US dollars, equivalent to about 8.36 billion yuan, across multiple rounds before its public listing, with a 2023 valuation reported near 2.8 billion US dollars. [7][19] Reported investors include large Chinese technology firms and funds such as Alibaba, Ant Group, Tencent, Meituan, and Xiaomi, the venture firm HongShan, and Middle Eastern backers including Saudi Aramco's Prosperity7 Ventures. [7][19] The firm has also drawn capital from Chinese state linked investors, including a financing from Shanghai state funds in 2025. [20]
In January 2025 the United States Department of Commerce added Zhipu AI and several affiliated entities to its export control Entity List, restricting their access to certain US technology, as part of a broader action against Chinese firms tied to advanced computing. [21][22] The company said it strongly disagreed with the decision, called the listing without factual basis, and stated that it did not expect a substantial impact on its operations. [21][22]
Zhipu listed on the Hong Kong Stock Exchange under stock code 2513 and began trading on 8 January 2026. [7][17] The offering raised roughly 558 million US dollars and valued the company at about 6.6 billion US dollars, and coverage described Zhipu as the first major Chinese generative AI company, and one of the first dedicated AGI focused foundation model companies in the world, to go public. [7][8][17]
Tang has been elected a Fellow of three major professional societies in computing and artificial intelligence. He was named an IEEE Fellow in 2021, an ACM Fellow in 2022, and an AAAI Fellow in 2023. [6] He received the ACM SIGKDD Test of Time Award in 2020 for the ArnetMiner work and has held the National Science Fund for Distinguished Young Scholars. [1][6][12] Following Zhipu AI's public listing, financial press reporting estimated his personal net worth in the billions of US dollars based on his stake in the company. [9]
| Field | Detail |
|---|---|
| Full name | Tang Jie (唐杰) |
| Born | 1977 |
| Nationality | Chinese |
| Fields | Data mining, social networks, knowledge graphs, large language models |
| Institution | Tsinghua University, Department of Computer Science and Technology |
| Research group | Knowledge Engineering Group (KEG) |
| PhD | Tsinghua University, 2006 |
| Known for | AMiner; GLM and ChatGLM models; co-founding Zhipu AI |
| Company | Zhipu AI (Z.ai), co-founder and chief scientist |
| Society fellowships | IEEE Fellow (2021), ACM Fellow (2022), AAAI Fellow (2023) |
| Major award | ACM SIGKDD Test of Time Award (2020) |