Xiaoyuan Liu
Portrait of Xiaoyuan Liu

Xiaoyuan Liu

刘效源

About Me

I am a Ph.D. student in Computer Science (Sep. 2024 - present) at School of Data Science, The Chinese University of Hong Kong, Shenzhen (CUHK-Shenzhen), where I am fortunate to be advised by Prof. Pinjia He. Prior to this, I received my B.Eng. degree in Computer Science and Engineering from CUHK-Shenzhen in 2024. I am currently an Ace intern at Xiaohongshu, mentored by Dr. Wenxiang Jiao.

My research focuses on autonomous, self-aware LLM agents and LLM reasoning.

News

Scroll for earlier updates

  • 1 paper is accepted by NeurIPS 2026 (Evaluations and Benchmarks) 🎉

  • 1 paper is accepted by EMNLP 2026 Findings 🎉

  • Our team will present the paper at ICLR 2026 in Rio de Janeiro, Brazil 🇧🇷

  • I was granted the Duan Yong Ping Travel Award AY2025-26 🎉

  • 1 paper is accepted by ACL 2026 Main Conference 🎉

  • 1 paper is accepted by ICLR 2026 Conference 🎉

  • I will present 2 posters at NeurIPS 2025 Conference in San Diego, U.S.A. 🇺🇸

  • We will attend ChinaSoft 2025 for discussion on AI for Software Engineering

  • 2 papers are accepted by NeurIPS 2025 Conference 🎉

  • 1 paper is accepted by EMNLP 2025 Main Conference 🎉

  • This personal website is launched!

  • I will present 2 posters at ACL Main Conference in Vienna, Austria 🇦🇹

  • 2 papers are accepted by ACL 2025 Main Conference 🎉

Selected Publications

All publications

First author

  1. Under Review

    miniFlux: Exploring Agent-Directed Harness Evolution for Online Adaptation

    Xiaoyuan Liu, Wenxiang Jiao, Yuan Lu, GUO YI, Pinjia He

  2. ICLR 2026

    The Pensieve Paradigm: Stateful Language Models Mastering Their Own Context

    Xiaoyuan Liu, Tian Liang, Dongyang Ma, Deyu Zhou, Haitao Mi, Pinjia He, Yan Wang

  3. NeurIPS 2025

    Trust, But Verify: A Self-Verification Approach to Reinforcement Learning with Verifiable Rewards

    Xiaoyuan Liu, Tian Liang, Zhiwei He, Jiahao Xu, Wenxuan Wang, Pinjia He, Zhaopeng Tu, Haitao Mi, Dong Yu

  4. ACL 2025

    Insight Over Sight: Exploring the Vision-Knowledge Conflicts in Multimodal LLMs

    Xiaoyuan Liu, Wenxuan Wang, Youliang Yuan, Jen-tse Huang, Qiuzhi Liu, Pinjia He, Zhaopeng Tu

Co-author

  1. ACL 2025

    Can't See the Forest for the Trees: Benchmarking Multimodal Safety Awareness for Multimodal LLMs

    Wenxuan Wang, Xiaoyuan Liu, Kuiyi Gao, Jen-tse Huang, Youliang Yuan, Pinjia He, Shuai Wang, Zhaopeng Tu

Teaching

CUHK-Shenzhen

Services

Program Comittee

  • ACM CIKM 2026 Workshop LASS

Conference reviewer

  • NeurIPS 2026
  • ICLR 2026, 2027
  • ACL Rolling Review (2025)

Journal reviewer

  • ACM Transactions on Software Engineering and Methodology (TOSEM)

Education

Work Experience

Honors & Awards