I am a third-year Master’s student in Software Engineering at Peking University, advised by Professor Li Weiping. To date, I have published 8 papers (as first author or co-first author) at top-tier venues including ACL, WWW, and CVPR. I am currently working at Xiaohongshu Pevek Foundation Model Team, where my research focuses on post-training and reinforcement learning for search foundation models.

Research Interests: My primary research areas include Large Language Models (LLMs) and Vision-Language Models (VLMs). Over the past year, I have been dedicated to enhancing the reasoning capabilities of LLMs and VLMs, as well as improving tool-calling ability and accuracy.

🔥 News

• 2026.4:  🎉🎉 We have 3 papers accepted to the ACL 2026 main conference.!

• 2026.2:  🎉🎉 Our paper “Residual Decoding: Mitigating Hallucinations in Large Vision-Language Models via History-Aware Residual Guidance” was accepted to CVPR 2026!

• 2026.1:  🎉🎉 Our paper “Accurate and Efficient Personalized Query Rewriting in Baidu Search” was accepted to WWW 2026!

📝 Selected Publications

• Towards Order Fairness: Mitigating LLMs Order Sensitivity through Dual Group Advantage Optimization

  Xu Chu, Guanyu Wang, Zhijie Tan, Xinrong Chen, Ziyu Li, Tong Mo, Weiping Li

  In Proc. of The 64rd Annual Meeting of the Association for Computational Linguistics (ACL).

• RADO: Reasoning Audit-Driven Optimization for Rigorous Reasoning in High-Stakes Domains

  Zhijie Tan*, Xu Chu*, Guanyu Wang, Ziyu Li, Weiping Li, Tong Mo

  In Proc. of The 64rd Annual Meeting of the Association for Computational Linguistics (ACL).

• MuSe: Multi-Stage Graph Reasoning via Vision-Language Models

  Guanyu Wang*, Xu Chu*, Zhijie Tan, Xinrong Chen, Tong Mo, Weiping Li

  In Proc. of The 64rd Annual Meeting of the Association for Computational Linguistics (ACL).

• Accurate and Efficient Personalized Query Rewriting in Baidu Search

  Xu Chu, Angela Li, Jiaming Zhang, Wei Li, Zhijie Tan, Dawei Yin, Shuaiqiang Wang, Daiting Shi

  In Proc. of The 35th International World Wide Web Conference (WWW).

• Residual Decoding: Mitigating Hallucinations in Large Vision-Language Models via History-Aware Residual Guidance

  Xinrong Chen*, Xu Chu*, Yingmin Qiu, Hengyuan Zhang, Jing Xiong, Shiyu Tang, Shuai Liu, Shaokang Yang, Cheng Yang, Hayden Kwok-Hay So, Ngai Wong

  In Proc. of The IEEE/CVF Conference on Computer Vision and Pattern Recognition 2026 (CVPR).

• Domain$o1$s: Guiding LLM Reasoning for Explainable Answers in High-Stakes Domains

  Xu Chu*, Zhijie Tan*, Hanlin Xue, Guanyu Wang, Tong Mo, Weiping Li

  In Proc. of The 63rd Annual Meeting of the Association for Computational Linguistics (ACL).

• Adaptive Spatiotemporal Augmentation for Improving Dynamic Graph Learning

  Xu Chu*, Hanlin Xue*, Bingce Wang, Xiaoyang Liu, Weiping Li, Tong Mo, Tuoyu Feng, Zhijie Tan

  In Proc. of ICASSP 2025 - 2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

✏️ Preprints

• Qwen Look Again: Guiding Vision-Language Reasoning Models to Re-attention Visual Information

  Xu Chu*, Xinrong Chen*, Guanyu Wang*, Zhijie Tan, Kui Huang, Wenyu Lv, Tong Mo, Weiping Li

• GraphSOS: Graph Sampling and Order Selection to Help LLMs Understand Graphs Better

  Xu Chu*, Hanlin Xue*, Zhijie Tan, Bingce Wang, Tong Mo, Weiping Li

• Order matters: Exploring order sensitivity in multimodal large language models

  Zhijie Tan*, Xu Chu*, Weiping Li, Tong Mo

📖 Educations

• 2023.09 – Present: Master’s student in Software Engineering under the supervision of Prof. Weiping Li at Peking University.
• 2019.09 – 2023.07: Bachelor of Engineering, School of Electronic Information, Wuhan University.

💻 Internships

• 2024.01 – 2024.05, NIO Shanghai, China.
  Contributed to the development of the LLM framework BI Agent, primarily responsible for intent understanding, task routing, and reasoning using LLMs.

• 2025.04 – 2026.01, Baidu, China.
  During the summer internship, I received the Xinghai Talent Program. I participated in the pre-training of Search LLM and the reinforcement learning post-training of Query Rewriting LLM. My main focus was on compressing LLM chain-of-thought to reduce the inference time of the query rewriting LLM.

• 2026.04 – present, Xiaohongshu, China.
  I received the offer of Xiaohongshu Pevek Foundation Model Team. Primarily responsible for research on post-training and reinforcement learning for search foundation models.

🗺️ Visitors