AI NEWS PAGE DOSSIER
Junyang Lin
Junyang Lin's file: Peking University, Alibaba DAMO, Qwen, open-weight model releases, billion-download scale, and the technical leader who walked away at the peak.
Lin joined Alibaba as a young language-model researcher, became one of the technical leaders behind Qwen's rapid open-model expansion, helped push the family from basic chat models into long-context and multimodal systems, then resigned in 2026 as the project reached enormous global distribution.
HOW THE HELL DID JUNYANG LIN GET HERE?
Jul 2019
HE JOINS ALIBABA'S AI RESEARCH MACHINE
Lin joined Alibaba in 2019 after graduate study and began working inside the company's language-model research organization.
RECEIPTS: TechCrunchreported
Apr 2023
HE MOVES ONTO THE NEW QWEN TEAM
Lin joined the Qwen effort in April 2023 as Alibaba organized a dedicated team around its new foundation-model family.
RECEIPTS: TechCrunchreported
Sep 28, 2023
QWEN GETS ITS FIRST TECHNICAL REPORT
Lin was among the authors of the first Qwen technical report covering base, chat, coding and mathematics variants of Alibaba's model family.
RECEIPTS: Qwenreported
Jul 15, 2024
QWEN2 EXPANDS THE OPEN MODEL FAMILY
Lin co-authored the Qwen2 technical report as Alibaba expanded the family across dense and mixture-of-experts model sizes.
RECEIPTS: Qwenreported
Jan 26, 2025
THE CONTEXT WINDOW STRETCHES TO A MILLION TOKENS
Lin co-authored the Qwen2.5-1M report describing a million-token context window and open inference techniques.
RECEIPTS: Qwenreported
Mar 4, 2026
THE QWEN TECH LEAD WALKS OUT
Lin resigned from Alibaba's Qwen organization after helping lead a model family that Reuters said had surpassed one billion downloads across hundreds of open releases.
RECEIPTS: Reutersreported