About Me

I am a second-year PhD candidate at HKUST NLP Group supervised by Professor Junxian He. I graduated from Shanghai Jiao Tong University (SJTU) in June 2024, where I was also fortunate to be advised by Prof. Junxian He during my undergraduate studies.

Research Interests

My research focuses on natural language processing and machine learning, with particular emphasis on:

• Agentic LLM & Long-Horizon Reasoning: Developing long-horizon reasoning web agents
• Logical Reasoning & Reinforcement Learning: Exploring reinforcement learning on verifiable logical reasoning data
• Vision-Language Models: Investigating perception bottlenecks and hallucination issues in LVLMs (How does ViT ability affect the LVLMs?)
• LLM Truthfulness & Interpretability: Understanding the internal mechanisms of truthfulness in large language models

Education

• Ph.D. in Computer Science (2024-Present)
  Hong Kong University of Science and Technology

• B.Eng. (2020-2024)
  Shanghai Jiao Tong University

Experience

• Research Intern (February 2025 - Present)
  MINIMAX

• Research Intern (June 2024 - September 2024)
  Tencent WXG
  Advisor: Zifei Shan

• Research Intern (June 2023 - December 2023)
  Shanghai AI Lab
  Advisor: Prof. Yu Cheng

Publications

2025

• WebExplorer: Explore and Evolve for Training Long-Horizon Web Agents
  Junteng Liu, Yunji Li, Chi Zhang, Jingyang Li, Aili Chen, Ke Ji, Weiyu Cheng, Zijia Wu, Chengyu Du, Qidi Xu, Jiayuan Song, Zhengmao Zhu, Wenhu Chen, Pengyu Zhao, Junxian He (Arxiv)
  Code

• SynLogic: Synthesizing Verifiable Reasoning Data at Scale for Learning Logical Reasoning and Beyond
  Junteng Liu, Yuanxiang Fan, Zhuo Jiang, Han Ding, Yongyi Hu, Chi Zhang, Yiqi Shi, Shitong Weng, Aili Chen, Shiqi Chen, Yunan Huang, Mozhi Zhang, Pengyu Zhao, Junjie Yan, Junxian He (Arxiv)
  Code

• On the Perception Bottleneck of VLMs for Chart Understanding
  Junteng Liu, Weihao Zeng, Xiwen Zhang, Yijun Wang, Zifei Shan, Junxian He (EMNLP 2025 Findings)
  Code

2024

• On the Universal Truthfulness Hyperplane Inside LLMs
  Junteng Liu, Shiqi Chen, Yu Cheng, Junxian He
  Conference on Empirical Methods in Natural Language Processing (EMNLP 2024)
  Code

• In-Context Sharpness as Alerts: An Inner Representation Perspective for Hallucination Mitigation
  Shiqi Chen, Miao Xiong, Junteng Liu, Zhengxuan Wu, Teng Xiao, Siyang Gao, Junxian He
  International Conference on Machine Learning (ICML 2024)

2023

• C-Eval: A Multi-Level Multi-Discipline Chinese Evaluation Suite for Foundation Models
  Yuzhen Huang, Yuzhuo Bai, Zhihao Zhu, Junlei Zhang, Jinghan Zhang, Tangjun Su, Junteng Liu, Chuancheng Lv, Yikai Zhang, Jiayi Lei, Yao Fu, Maosong Sun, Junxian He
  Neural Information Processing Systems (NeurIPS 2023)

• Composing Parameter-Efficient Modules with Arithmetic Operations
  Jinghan Zhang, Shiqi Chen, Junteng Liu, Junxian He
  Neural Information Processing Systems (NeurIPS 2023)

Honors & Awards

• Zhiyuan Honor Scholarship, Shanghai Jiao Tong University

Contact

• Email: [jliugi@connect.ust.hk]
• GitHub: Vicent0205