
Yuze Zhao
Fifth-Year PhD Candidate in Computer Science
About Me
I am Yuze Zhao, a PhD candidate in Computer Science at the University of Science and Technology of China. I conduct research on large language models at the State Key Laboratory of Cognitive Intelligence and am a member of the BASE Group. I am advised by Prof. Qi Liu and Asst. Prof. Zhenya Huang.
My research interests include code reasoning, agentic reinforcement learning, human-computer interaction, and trustworthy large language model systems.
I am currently an intern at Qwen Token Foundry, working on Cowork RL: reinforcement learning for agents that plan, use tools, and carry out complex tasks in real-world workflows.
Bio
News
Recent research, experience, and honor updates.
Education
Experience
Publications
Selected papers are listed first, followed by other accepted papers and preprints.
The Forty-third International Conference on Machine Learning (ICML 2026).
@inproceedings{zhao2026what,
title={What Really Improves Mathematical Reasoning: Structured Reasoning Signals Beyond Pure Code},
author={Zhao, Yuze and Fang, Junpeng and Yu, Lu and Huang, Zhenya and Zhang, Kai and Cui, Qing and Liu, Qi and Zhou, Jun and Chen, Enhong},
booktitle={Forty-third International Conference on Machine Learning},
year={2026},
url={https://openreview.net/forum?id=7EK1E8wxVO}
}
The Thirteenth International Conference on Learning Representations (ICLR 2025), Singapore, 2025.
Introduces code reasoning as a distinct reasoning setting and proposes the RHDA pipeline to iteratively decompose, verify, and amend hypotheses during program reasoning.
@inproceedings{zhao2025unveiling,
title={Unveiling the Magic of Code Reasoning through Hypothesis Decomposition and Amendment},
author={Zhao, Yuze and Ji, Tianyun and Feng, Wenjun and Huang, Zhenya and Liu, Qi and Liu, Zhiding and Ma, Yixiao and Zhang, Kai and Chen, Enhong},
booktitle={The Thirteenth International Conference on Learning Representations},
year={2025},
url={https://openreview.net/forum?id=kN25ggeq1J}
}
Findings of the Association for Computational Linguistics: ACL 2024, pages 16415-16429, Bangkok, Thailand. Association for Computational Linguistics.
@inproceedings{zhao2024repair,
title={RePair: Automated Program Repair with Process-based Feedback},
author={Zhao, Yuze and Huang, Zhenya and Ma, Yixiao and Li, Rui and Zhang, Kai and Jiang, Hao and Liu, Qi and Zhu, Linbo and Su, Yu},
booktitle={Findings of the Association for Computational Linguistics: ACL 2024},
pages={16415--16429},
year={2024},
url={https://aclanthology.org/2024.findings-acl.973/}
}
Semantic-Aligned Code Summarization: Bridging the Gap Between Code and Natural Language Through Data Flow Analysis
IEEE Transactions on Neural Networks and Learning Systems, 36(10): 19240-19254, 2025.
@article{zhao2025semantic,
title={Semantic-Aligned Code Summarization: Bridging the Gap Between Code and Natural Language Through Data Flow Analysis},
author={Zhao, Yuze and Huang, Zhenya and Zhang, Kai and Gao, Weibo and Liu, Qi and Liu, Xukai and Yao, Fangzhou and Chen, Enhong},
journal={IEEE Transactions on Neural Networks and Learning Systems},
volume={36},
number={10},
pages={19240--19254},
year={2025}
}
Findings of the Association for Computational Linguistics: ACL 2026, pages 39651-39674, San Diego, California, United States.
@inproceedings{sun-etal-2026-swe,
title={SWE-Mutation: Can LLMs Generate Reliable Test Suites in Software Engineering?},
author={Sun, Yuxuan and Zhao, Yuze and Wang, Yufeng and Du, Yao and Ma, Zhiyuan and Wang, Jinbo and Zhang, Mengdi and Zhang, Kai and Huang, Zhenya},
booktitle={Findings of the Association for Computational Linguistics: ACL 2026},
pages={39651--39674},
year={2026},
doi={10.18653/v1/2026.findings-acl.1976},
url={https://aclanthology.org/2026.findings-acl.1976/}
}
econvRBP: Improved Ensemble Convolutional Neural Networks for RNA Binding Protein Prediction Directly from Sequence
Methods, 181-182: 15-23, 2020.
@article{zhao2020econvrbp,
title={econvRBP: Improved Ensemble Convolutional Neural Networks for RNA Binding Protein Prediction Directly from Sequence},
author={Zhao, Yuze and Du, Xiuquan},
journal={Methods},
volume={181-182},
pages={15--23},
year={2020}
}
Proceedings of the AAAI Conference on Artificial Intelligence, 39(23): 24276-24284, 2025.
@inproceedings{jiang2025verse,
title={VERSE: Verification-based Self-Play for Code Instructions},
author={Jiang, Hao and Liu, Qi and Li, Rui and Zhao, Yuze and Ma, Yixiao and Ye, Shengyu and Lu, Junyu and Su, Yu},
booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
volume={39},
number={23},
pages={24276--24284},
year={2025}
}
Proceedings of the 31st ACM SIGKDD Conference on Knowledge Discovery and Data Mining, pages 354-365, 2025.
@inproceedings{gao2025denoising,
title={Denoising Programming Knowledge Tracing with a Code Graph-based Tuning Adaptor},
author={Gao, Weibo and Liu, Qi and Li, Rui and Zhao, Yuze and Wang, Hao and Yue, Linan and Yao, Fangzhou and Zhang, Zheng},
booktitle={Proceedings of the 31st ACM SIGKDD Conference on Knowledge Discovery and Data Mining},
pages={354--365},
year={2025}
}
Proceedings of the AAAI Conference on Artificial Intelligence, 38(8): 8679-8687, 2024.
@inproceedings{li2024consider,
title={CONSIDER: Commonalities and Specialties Driven Multilingual Code Retrieval Framework},
author={Li, Rui and He, Liyang and Liu, Qi and Zhao, Yuze and Zhang, Zheng and Huang, Zhenya and Su, Yu and Wang, Shijin},
booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
volume={38},
number={8},
pages={8679--8687},
year={2024}
}
Proceedings of the ACM Web Conference 2026, pages 5568-5578, Dubai, United Arab Emirates.
@inproceedings{pan2026adaptive,
title={Adaptive Model and Strategy Routing for Cost-Efficient LLM Services},
author={Pan, Zhihong and Zhang, Kai and Zhao, Yuze and Han, Yupeng},
booktitle={Proceedings of the ACM Web Conference 2026},
pages={5568--5578},
year={2026},
publisher={ACM},
doi={10.1145/3774904.3792556}
}DisenTS: Disentangled Channel Evolving Pattern Modeling for Multivariate Time Series Forecasting
IEEE Transactions on Pattern Analysis and Machine Intelligence, 48(10): 11808-11821, 2026.
@article{liu2026disents,
title={DisenTS: Disentangled Channel Evolving Pattern Modeling for Multivariate Time Series Forecasting},
author={Liu, Zhiding and Yang, Jiqian and Mao, Qingyang and Zhao, Yuze and Cheng, Mingyue and Li, Zhi and Liu, Qi and Chen, Enhong},
journal={IEEE Transactions on Pattern Analysis and Machine Intelligence},
volume={48},
number={10},
pages={11808--11821},
year={2026},
doi={10.1109/TPAMI.2026.3697127}
}Frontiers of Computer Science, 21(4): 2104704, 2027. Published online in 2026.
@article{ma2027position,
title={Position-aware modeling for fine-grained visually-rich long document understanding},
author={Ma, Yixiao and Ruan, Shulan and Song, Zijie and Zhang, Xin and Zhao, Yuze and Huang, Zhenya and Chen, Enhong},
journal={Frontiers of Computer Science},
volume={21},
number={4},
pages={2104704},
year={2027},
doi={10.1007/s11704-026-51131-x}
}Honors
Scholarships, awards, and competition results.