我是何欣程,本科与博士(推免直博)均毕业于南京大学计算机学院,导师为徐宝文教授,博士期间研究软件质量保证与动态语言 API 缺陷管理。毕业后曾在企业从事 AI for Software Engineering 前沿技术研究。现为江南大学人工智能与计算机学院校聘副教授,研究聚焦 SE4AI:智能体(Agent)的可信与安全。欢迎通过邮件交流学术问题。
I am Xincheng He. I received my Bachelor's and Ph.D. degrees (direct-entry Ph.D. program) from the School of Computer Science, Nanjing University, supervised by Prof. Baowen Xu. My doctoral research focused on software quality assurance and API misuse management for dynamic languages. After graduation, I conducted industry research on AI for Software Engineering. I am currently an Associate Professor at the School of Artificial Intelligence and Computer Science, Jiangnan University, focusing on SE4AI: trustworthiness and security of LLM agents. Feel free to reach out by email for academic discussions.
我的研究以程序分析与软件质量为根基,向 AI 智能软件工程与智能体可信安全延伸,当前沿三个方向展开:My research is rooted in program analysis and software quality, extending toward AI-powered software engineering and trustworthy agents. It currently unfolds along three directions:
利用大模型与 Agent 技术赋能软件工程任务:智能编程辅助、代码质量看护、测试自动化与缺陷检测。Leveraging LLMs and agent techniques to empower software engineering tasks: intelligent programming assistance, code quality monitoring, test automation, and defect detection.
面向 LLM 工具调用与智能体系统的静默语义异常检测与归因,构建可解释、可信赖的智能体软件系统。Detecting and attributing silent semantic anomalies in LLM tool calls and agent systems, toward explainable and trustworthy agent software.
静态/混合程序分析、API 误用检测与推荐、二进制函数名恢复与逆向工程、安卓分析与 UI 测试。Static/hybrid program analysis, API misuse detection and recommendation, binary function-name recovery and reverse engineering, Android analysis and UI testing.
已发表/录用学术论文 10 篇,授权国家发明专利 5 项。全部论文如下(粗体为本人在作者列表中):📕CCF-A · 📘CCF-B/C · 📗期刊/其他10 accepted/published papers and 5 granted Chinese patents. All papers are listed below (bold marks my name):📕CCF-A · 📘CCF-B/C · 📗journal/other
🏛️ 江南大学 · 人工智能与计算机学院,无锡School of Artificial Intelligence and Computer Science, Jiangnan University, Wuxi
✉️ 邮箱Email: xche@jiangnan.edu.cn
🔗 GitHub: github.com/njuxc
欢迎对 AI4SE、Agent 可信与安全、程序分析感兴趣的同行与学生邮件交流。Researchers and students interested in AI4SE, trustworthy agents, or program analysis are welcome to reach out by email.