Scholar
Chujie Zheng
Google Scholar ID: 55zBNgUAAAAJ
Qwen Team, Alibaba Group
Citations & Impact
Citations
6,608
H-index
26
i10-index
29
Publications
20
Co-authors
6
list available
Publications
Resume (English only)
Academic Achievements
- Selected for the 2025 CIPS Doctoral Dissertation Incentive Program
- Papers '80/20 Rule' and 'SuperGPQA' accepted to NeurIPS 2025
- 'ProcessBench' paper awarded the ACL 2025 SAC Award
- Recipient of the 2025 WAIC Yunfan Award (Spotlight, 'Rising Star')
- Led or contributed to the release of key models and reports: Qwen3 series (e.g., Qwen3-235B, Qwen3-Coder-480B, Qwen3-Thinking), QwQ-32B reasoning model, Yi-Lightning
- Released multiple benchmarks and algorithms: SuperGPQA (comprehensive LLM evaluation), ProcessBench (process supervision for math reasoning), GSPO (large-scale RL), WorldPM (world preference modeling), ExPO (efficient alignment), DRO (safety prompt optimization), PriDe (debiasing in multiple-choice QA)
- Published papers at top-tier conferences: ACL 2025 (ExPO, ProcessBench, Qwen2.5-Math-PRM), ICML 2024 (DRO), ICLR 2024 Spotlight (PriDe, top 5%)
Background
- Researcher at the Qwen Team, Alibaba Group
- Dedicated to building scalable, generalist AI systems
- Focuses on methodologies that consistently and efficiently improve AI intelligence and task-solving abilities with increased compute and data
- Current work centers on advancing reasoning capabilities of Qwen models (e.g., Qwen3, QwQ) and developing large-scale reinforcement learning (RL) approaches
- Broad research interests include model architecture, interpretability, safety, and alignment
- Previously conducted extensive research on LLMs for social good, particularly emotional support systems
