PeopleofSuperintelligence

Lab leader

Xin Jiang

Xin Jiang works on large language models at Huawei. He completed a Ph.D. in applied mathematics at Peking University in 2009, after a B.S. in theoretical and applied mechanics, and then worked as a senior engineer at NetEase Youdao and Baidu. He joined Huawei in 2013, directed the Speech and Semantics Lab at Noah's Ark Lab from 2017 to 2023, and from 2020 to early 2025 was one of the technical leaders for PanGu-Alpha and PanGu-Sigma on Ascend NPUs. In 2020, he coauthored TinyBERT, and in 2022 a paper he coauthored on compression of generative pre-trained language models via quantization received an ACL Outstanding Paper award. Since March 2025, he is a senior expert at the AI Lab of Huawei Hong Kong Research Center.

Now at
Huawei and Huawei Hong Kong Research Center
Works
176
Citations
6,790
h-index
36

Publication numbers from OpenAlex, as of 17 October 2026.

Timeline

  1. 2000–2009

    Applied mathematics at Peking

    He earns a B.S. in theoretical and applied mechanics at Peking University from 2000 to 2004 and a Ph.D. in applied mathematics there from 2004 to 2009, supervised by Ming Jiang, Bin Yu, and Hang Li.

  2. July 2009

    A ranking approach to keyphrases

    In July 2009, he, Yunhua Hu, and Hang Li publish a ranking approach to keyphrase extraction that uses Ranking SVM.

  3. 2009–2013

    Search engineering in Beijing

    He works as a senior engineer at NetEase Youdao from 2009 to 2011 and at Baidu from 2011 to 2013.

  4. 2013–February 2025

    PanGu at Noah's Ark Lab

    He joins Huawei in 2013, directs the Speech and Semantics Lab at Noah's Ark Lab from 2017 to 2023, and serves as principal researcher there from November 2023 to February 2025. From 2020 to early 2025, he is one of the technical leaders for PanGu-Alpha, a 200 billion-parameter autoregressive model, and PanGu-Sigma, a trillion-parameter sparse MoE model, on Ascend NPUs.

  5. November 2020

    Distilling BERT into TinyBERT

    In November 2020, he coauthors TinyBERT, which reaches more than 96.8% of BERT-Base performance on GLUE while being 7.5 times smaller and 9.4 times faster at inference.

  6. 2022

    An ACL Outstanding Paper

    In 2022, his work on compression of generative pre-trained language models via quantization receives an ACL Outstanding Paper award.

  7. March 2023

    PanGu-Sigma at trillion scale

    In March 2023, PanGu-Σ is presented with 1.085 trillion parameters, trained on Ascend 910 processors, and Xin Jiang is among the authors.

  8. Since March 2025

    Senior expert in Hong Kong

    Since March 2025, he is a senior expert at the AI Lab of Huawei Hong Kong Research Center and manager of the large-model team in the Small and Agile Commando Team, working on large language models.

Sources and credits

Photo: We have no published photo of Xin Jiang yet. Send us a free-licensed photo.

  1. Xin Jiang jxfeb.github.io. Checked 16 October 2026.
  2. A Ranking Approach to Keyphrase Extraction - Microsoft Research microsoft.com. Checked 16 October 2026.
  3. TinyBERT: Distilling BERT for Natural Language Understanding - ACL Anthology aclanthology.org. Checked 16 October 2026.
  4. Compression of Generative Pre-trained Language Models via Quantization - ACL Anthology aclanthology.org. Checked 16 October 2026.
  5. PanGu-Σ: Towards Trillion Parameter Language Model with Sparse Heterogeneous Computing arxiv.org. Checked 16 October 2026.