I am a master’s student in Software Engineering at Zhejiang University, advised by Dr. Tianyu Du and Prof. Shouling Ji. Before that, I received my bachelor’s degree in Cyberspace Security from Nanjing University of Science and Technology.

My research focuses on LLM post-training, reliable agents, and AI safety. I am particularly interested in enhancing the reasoning and agentic capabilities of large language models through post-training and reinforcement learning, while improving their safety and reliability in real-world applications.

I have six first-author papers accepted at NeurIPS, ICLR, AAAI, WWW, ACL, and EMNLP. During my master’s studies, I have also had the opportunity to collaborate with researchers from industry, including Taobao & Tmall Group of Alibaba and OPPO Research Institute.

I am currently open to opportunities in LLM algorithm engineering. Feel free to get in touch.

🔥 News

• 2026.09:   🎉 One paper was accepted by NeurIPS 2026.
• 2026.04:   🎉 One paper was accepted by ACL 2026.
• 2026.01:   🎉 One paper was accepted by ICLR 2026.
• 2026.01:   🎉 One paper was accepted by WWW 2026.
• 2025.11:   🎉 One paper was selected for ORAL presentation at AAAI’26!
• 2025.11:   🎉 One paper was accepted by AAAI 2026.
• 2025.08:   🎉 One paper was accepted by EMNLP 2025.

📝 Publications

NeurIPS 2026

StateTree: Enhancing Long-term Dialogue Reasoning via Reinforcement Learning

Naen Xu*, Wanqing Cui*, Yibo Hu, Shixin Hong, Hengyu An, Meiguang Jin, Junfeng Ma, Tianyu Du, NeurIPS 2026. [Code] [CCF A]

• A state-tree reasoning task that teaches models to retrieve, update, and combine information across long conversations through reinforcement learning. StateTree-14B generalizes from 10K-context training to 128K dialogue evaluation, reaching 74.2 F1 on LoCoMo.

• NeurIPS 2026 StateTree: Enhancing Long-term Dialogue Reasoning via Reinforcement Learning, Naen Xu*, Wanqing Cui*, Yibo Hu, Shixin Hong, Hengyu An, Meiguang Jin, Junfeng Ma, Tianyu Du, NeurIPS 2026. [Code] [CCF A]
• ICLR 2026 When Agents “Misremember” Collectively: Exploring the Mandela Effect in LLM-based Multi-Agent Systems, Naen Xu, Hengyu An, Shuo Shi, Jinghuai Zhang, Chunyi Zhou, Changjiang Li, Tianyu Du, Zhihui Fu, Jun Wang, Shouling Ji, ICLR 2026. [Code] [CCF A]
• WWW 2026 FraudShield: Knowledge Graph Empowered Defense for LLMs against Fraud Attacks, Naen Xu*, Jinghuai Zhang*, Ping He, Chunyi Zhou, Jun Wang, Zhihui Fu, Tianyu Du, Zhaoxiang Wang, Shouling Ji, WWW 2026. [Code] [CCF A]
• AAAI 2026 Bridging the Copyright Gap: Do Large Vision-Language Models Recognize and Respect Copyrighted Content?, Naen Xu*, Jinghuai Zhang*, Changjiang Li, Hengyu An, Chunyi Zhou, Jun Wang, Boyu Xu, Yuyuan Li, Tianyu Du, Shouling Ji, AAAI (Oral) 2026. [Code] [CCF A]
• ACL 2026 “I See What You Did There”: Can Large Vision-Language Models Understand Multimodal Puns?, Naen Xu, Jiayi Sheng, Changjiang Li, Chunyi Zhou, Yuyuan Li, Jun Wang, Zhihui Fu, Tianyu Du, Jinbao Li, Shouling Ji, ACL Main 2026. [Code] [CCF A]
• EMNLP 2025 VideoEraser: Concept Erasure in Text-to-Video Diffusion Models, Naen Xu*, Jinghuai Zhang*, Changjiang Li, Zhi Chen, Chunyi Zhou, Qingming Li, Tianyu Du, Shouling Ji, EMNLP (Main) 2025. [Code] [TH-CPL A]

🎓 Education

• 2024.09 – 2027.06 (expected), M.E., Software Engineering, Zhejiang University, Hangzhou.
• 2020.09 - 2024.06, B.E., Cyberspace Security, Nanjing University of Science and Technology, Nanjing.

🎖 Honors and Awards

• 2026.10 National Scholarship (Master)
• 2023.10 National Scholarship (Undergraduate)