Hi, I am Chunkang Zhang (张淳慷).
🎓 I obtained my master’s degree from the University of Chinese Academy of Sciences (UCAS), advised by Prof. Le Sun and Prof. Yaojie Lu.
🔬 My research interests include Agentic Memory, Retrieval-Augmented Generation (RAG), Long-Context Comprehension, and Omni modal LLM. I work closely with Jialong Tang and Chulun Zhou at Qwen, Alibaba. I also enjoyed close collaboration with Mo Yu at Hunyuan, Tencent.
💡 I believe that true intelligence requires not just reasoning, but remembering — forming structured, evolving knowledge from raw experience.
📩 If you are interested in my work or potential collaboration, feel free to email me at zkang5051@gmail.com.
🔥 News
- [2026.05] 🎉🎉 HGMem is accepted at ICML 2026!
- [2026.05] 🎉🎉 Self-Jailbreak is accepted at ACL 2026 Findings！
- [2026.01] 🏅 HGMem is featured as #1 Paper of the Day on Hugging Face!
- [2025.12] 📄 HGMem is released on arXiv — a hypergraph-based working memory mechanism for multi-step RAG.
- [2025.10] 📄 Self-Jailbreak is released on arXiv — unveiling and mitigating self-jailbreak risks in Large Reasoning Models.
- [2025.09] 🚀 The LongCat-Flash technical report is released by the Meituan LongCat team — a 560B MoE model trained on 20T+ tokens.
- [2025.08] 🎉🎉 AUTOALIGN is accepted at ACL 2025.
- [2025.03] 🎉🎉 Our paper on the influence of external information on LLMs is published in IEEE TCSS 2025.
- [2024.10] Concluded internship at Meituan LLM Group after contributing to the LongCat series.
- [2023.08] 🚀 Started internship at Meituan LLM Group, working on long-context LLMs.
📝 Publications
Chulun Zhou*, Chunkang Zhang*, Guoxin Yu, Fandong Meng, Jie Zhou, Wai Lam, Mo Yu
🏅 #1 Paper of the Day on Hugging Face | arXiv | GitHub | HuggingFace
ACL 2026 Findings When Models Outthink Their Safety: Unveiling and Mitigating Self-Jailbreak in Large Reasoning Models
Yingzhi Mao*, Chunkang Zhang*, Junxiang Wang, Xinyan Guan, Boxi Cao, Yaojie Lu, et al.
ACL 2025 AUTOALIGN: Get Your LLM Aligned with Minimal Annotations
Xinyu Lu*, Dong Xu*, Chunkang Zhang*, Xinyan Guan, Junxiang Wang, Qingyu Zhang, et al.
Under Review Investigating Reasoning Models in Retrieval-Augmented Generation
Xinyan Guan, Jiali Zeng, Fandong Meng, Chunkang Zhang, Yaojie Lu, Hongyu Lin, et al.
Under Review ChameleonInstruct: Teaching Machines to Reason via Chameleon Game
Chunkang Zhang, Yaojie Lu, Hongyu Lin, Cao Liu, Xinyan Guan, Ke Zeng, Guanglu Wan, et al.
IEEE TCSS 2025 Influence of External Information on Large Language Models Mirrors Social Cognitive Patterns
Ning Bian, Hongyu Lin, Peilin Liu, Yaojie Lu, Chunkang Zhang, Ben He, Xianpei Han, Le Sun
Chunkang Zhang, Boxi Cao, Yaojie Lu, Hongyu Lin, Cao Liu, Ke Zeng, Guanglu Wan, et al.
ICAICA 2021 A Non-Autoregressive Network for Chinese Text to Speech and Voice Cloning
Chunkang Zhang, Yueqing Cai, Wenbi Rao
💼 Research Experience
📖 Educations
- 2022.08 – 2025.06, M.Sc., Computer Science and Technology, University of Chinese Academy of Sciences (UCAS). Advisor: Prof. Le Sun.
- 2018.08 – 2022.06, B.S., Software Engineering (Elite Class), Wuhan University of Technology (WHUT). GPA: 3.9/4.0 (91.14/100).
🎖 Honors and Awards
- Academic Scholarship, UCAS
- Outstanding Thesis & Outstanding Graduate, WHUT
- Academic Scholarship & Outstanding Student Merit, WHUT
- Second Prize, Chinese Collegiate Computing Competition (National Level)
- Best Individual, Cambridge University Summer Exchange Program
🎨 Hobbies
Outside of research, I enjoy playing the piano 🎹 (Grade 10 Certification), staying active through fitness 🏋️, badminton 🏸, tennis 🎾, and keeping up with good books 📚. If you share any of these interests, feel free to reach out — always happy to connect!