Junteng Liu

Ph.D. Candidate at HKUST NLP Group

prof_pic.jpg

Hong Kong University of Science and Technology

Clear Water Bay, Hong Kong

I am a second-year Ph.D. candidate at HKUST NLP Group, advised by Professor Junxian He. I received my B.Eng. from Shanghai Jiao Tong University (SJTU) in 2024.

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

  • Agentic LLMs and long-horizon reasoning
  • Logical reasoning and reinforcement learning
  • Vision-language models and perception bottlenecks
  • LLM truthfulness and interpretability

I am currently a research intern at Apple MLR (Cupertino), mentored by Yizhe Zhang. Previously, I interned at MiniMax (contributing to the M1 and M2.x models), Tencent WXG, and Shanghai AI Lab.

I am broadly interested in building trustworthy and capable reasoning systems.

Google Scholar @junteng88716710

News

Apr 2026 Started a research internship at Apple MLR in Cupertino, mentored by Yizhe Zhang.
Oct 2025 Our paper The Tool Decathlon was released, benchmarking language agents for diverse and long-horizon task execution.
Sep 2025 Our paper WebExplorer was released on arXiv for long-horizon web-agent training.
May 2025 SynLogic was accepted to NeurIPS 2025.
May 2025 Our paper On the Perception Bottleneck of VLMs for Chart Understanding was accepted to EMNLP 2025 Findings.
Feb 2025 Started a research internship at MiniMax, working on large-scale logical reasoning data synthesis and training for the M1 and M2.x models.
Dec 2024 Our paper On the Universal Truthfulness Hyperplane Inside LLMs was presented at EMNLP 2024.
Oct 2024 Joined HKUST NLP Group as a Ph.D. student advised by Prof. Junxian He.
Jun 2024 Graduated from Shanghai Jiao Tong University with a B.Eng. in Automation (IEEE Honor Class).