Scholar
Zexiong Ma
Google Scholar ID: jmZifrkAAAAJ
Peking University
Citations & Impact
Citations
272
H-index
5
i10-index
4
Publications
9
Co-authors
6
list available
Publications
Resume (English only)
Academic Achievements
- - 2025.10: Three papers accepted by ICSE 2026
- - 2025.05: One paper accepted by ACL 2025
- - [ICSE 2026] Tool-integrated Reinforcement Learning for Repo Deep Search
- - [ICSE 2026] An Empirical Study on LLM-based Agents for Automated Bug Fixing
- - [ICSE 2026] Evaluating Generated Commit Messages with Large Language Models
- - [ACL 2025] SoRFT: Issue Resolving with Subtask-oriented Reinforced Fine-Tuning
- - [LCFM@ICML 2025] Enhancing Retrieval-Augmented Generation with Dehallucinating Parallel Context Extension
- - [NeurIPS 2024] Make Your LLM Fully Utilize the Context
- - [EMNLP 2024] Can LLMs Learn From Mistakes? An Empirical Study on Reasoning Tasks
- - [ICPC 2024] Compositional API Recommendation for Library-Oriented Code Generation
- - [preprint] Repository Structure-Aware Training Makes SLMs Better Issue Resolver
Research Experience
- - 2024.10 - present, Intern at Trae/Marscode Research, ByteDance, China. Working with Chao Peng, Pengfei Gao, and Xiangxin Meng
- - 2023.03 - 2024.09, Intern at DKI Group, Microsoft, China. Working with Zeqi Lin and Shengnan An
Education
- - 2021.09 - 2026.07 (expected), Peking University, Ph.D. in Computer Science, Adviser: Prof. Bing Xie
- - 2017.09 - 2021.07, Tianjin University, B.S. Degree in Computer Science
Background
- - Research Interests: Application of large language models in coding and reasoning tasks, including code generation, mathematical reasoning, long-context reasoning, etc.
- - Professional Field: Software Engineering, Computer Science
- - Brief Introduction: A final-year Ph.D. candidate at the Software Engineering Institute, Peking University, supervised by Prof. Bing Xie. Focused on rule-based reinforcement learning to enhance the reasoning and coding capabilities of LLMs.
Miscellany
- - Personal Interests: Not provided
