showSidebars ==
showTitleBreadcrumbs == 1
node.field_disable_title_breadcrumbs.value ==

Pre-Conference Talk by YU Jiongchi | Defects4C: Benchmarking Large Language Model Repair Capability with C/C++ Bugs

Please click here if you are unable to view this page.

 


Defects4C: Benchmarking Large Language Model Repair Capability with C/C++ Bugs
 

Speaker (s):


YU Jiongchi
PhD Candidate
School of Computing and Information Systems
Singapore Management University

Date:

Time:

Venue:

 

11 November 2025, Tuesday

3:05pm – 3:15pm

Meeting room 5.1, Level 5
School of Computing and
Information Systems 1, 
Singapore Management University, 
80 Stamford Road,
Singapore 178902

We look forward to seeing you at this research seminar.

Please register by 9 November 2025.

About the Talk

Automated Program Repair (APR) has achieved remarkable progress in Java, yet research on C/C++ repair remains limited due to the lack of suitable benchmarks. To bridge this gap, we present Defects4C, a comprehensive benchmark for C/C++ program repair built from real-world repositories. Defects4C provides millions of bug-related commits and hundreds of curated buggy and vulnerable functions with executable test cases, enabling systematic evaluation and model retraining. Leveraging this dataset, we evaluate 24 state-of-the-art foundation language models on C/C++ repair tasks, revealing key strengths, weaknesses, and future directions for LLM-based APR in various programming languages.

This is a Pre-Conference talk for 40th IEEE/ACM International Conference on Automated Software Engineering (ASE 2025).

About the speaker

Jiongchi YU is a Ph.D. candidate in Computer Science at the SMU School of Computing and Information Systems, supervised by Prof. Xiaofei XIE. His research interests include program analysis, software engineering and security.