About
I am currently a Postdoctoral Researcher at Zhejiang University, working with Prof. Xin Xia and Prof. Xing Hu.
Prior to this, I received my B.Sc. and M.Sc. degrees from Nantong University, supervised by Prof. Xiang Chen. I obtained my Ph.D. degree from Nanjing University of Aeronautics and Astronautics, supervised by Prof. Yu Zhou and Prof. Taolue Chen. From August 2024 to July 2025, I was a visiting scholar at Singapore Management University, advised by Prof. David Lo.
My current research focuses on Code Generation, a trending intersection of Artificial Intelligence and Software Engineering. I have published about 40 papers in total, including 5 first-authored CCF-A papers.
I am open to research discussions and collaborations. Please feel free to contact me.
Selected Publications
View All →Code-DiTing: Automatic Evaluation of Code Generation without References or Test Cases
Guang Yang, Yu Zhou†, Xiang Chen, Wei Zheng, Xing Hu, Xin Zhou, David Lo, Taolue Chen†
Proceedings of the 40th IEEE/ACM International Conference on Automated Software Engineering
A novel LLM-as-Judge method for code evaluation, balancing accuracy, efficiency and explainability.
Less is More: DocString Compression in Code Generation
Guang Yang, Yu Zhou†, Wei Cheng, Xiangyu Zhang, Xiang Chen, Terry Yue Zhuo, Ke Liu, Xin Zhou, David Lo, Taolue Chen†
ACM Transactions on Software Engineering and Methodology
A novel DocString compression method for code generation, achieving significant reduction in token processing cost while preserving the quality of the generated code.
Defending Code Language Models against Backdoor Attacks with Deceptive Cross-Entropy Loss
Guang Yang, Yu Zhou†, Xiangyu Zhang, Xiang Chen, Terry Zhuo, David Lo, Taolue Chen†
ACM Transactions on Software Engineering and Methodology
A general and effective loss function DeCE (Deceptive Cross-Entropy) to defend Code Language Models against backdoor attacks, preventing overfitting to backdoor triggers.
Chain-of-Thought in Neural Code Generation: From and for Lightweight Language Models
Guang Yang, Yu Zhou†, Xiang Chen, Xiangyu Zhang, Terry Yue Zhuo, Taolue Chen†
IEEE Transactions on Software Engineering
A novel approach COTTON to automatically generate CoTs for code generation, achieving significant improvement in performance compared to the state-of-the-art baselines.
How important are good method names in neural code generation? a model robustness perspective
Guang Yang, Yu Zhou†, Wenhua Yang, Tao Yue, Xiang Chen, Taolue Chen†
ACM Transactions on Software Engineering and Methodology
A novel approach RADAR to enhance the performance of PCGMs from a model robustness perspective.
News
One paper has been accepted by IPM!
Started working as Postdoc at Zhejiang University.
Completed and received my doctoral degree 🎉
Three papers have been accepted by ASE 2025!
One paper has been accepted by ACL Main!
One paper has been accepted by TSE!
Two papers have been accepted by TOSEM!