Kuan Zhang(张宽)

:blush: About Me

I am a senior undergraduate student at Beijing Institute of Technology (BIT), majoring in software engineering, while I am a coming PhD student from College AI of Tsinghua university.

My early research focused on data-centric AI, especially label noise and data selection. As vision-language models (VLMs) and reasoning models have grown impressively powerful, I’ve become deeply interested in the study of visual agents, especially utilizing VLMs in game area. My goal is to design general intelligent agents outperforming humans across all tasks in cyberspace.

In life, I have a deep appreciation for (e)sports, AAA game, music, novel and amination. I got a master degree in League of Legends.

📖 Academic Experience

  • 2026 - future, College AI, Tsinghua University
  • 2022 - 2026(expected), School of Computer Science and Technology, Beijing Institute of Technology
  • 2019 - 2022, Chengdu No.7 High School

📝 Publications

Handling Label Noise via Instance-Level Difficulty Modeling and Dynamic Optimization

  • NeurIPS 2025 Poster

  • Kuan Zhang, Chengliang Chai, Jingzhe Xu, Chi Zhang, Han Han, Ye Yuan, Guoren Wang, Lei Cao

  • An efficient, hyperparameter-free, instance-level optimization framework for label noise in image classification.

  • Paper Link

Harnessing Diversity for Important Data Selection in Pretraining Large Language Models

  • ICLR 2025 Spotlight

  • Chi Zhang, Huaping Zhong, Kuan Zhang, Chengliang Chai, Rui Wang, Xinlin Zhuang, Tianyi Bai, Qiu Jiantao, Lei Cao, Ju Fan, Ye Yuan, Guoren Wang, Conghui He

  • Combine multi-arm bandit with influence function to select data for pretraining process of large language models.

  • Paper Link

🏆 Selected Awards

2023/2024 National Scholarship (Highest honor for Chinese undergraduates)

2024 MCM M Prize

2023 CUMCM Second Prize

2023 CMC Second Prize