Papers by Jin L.c. Guo
DocChecker: Bootstrapping Code Large Language Model for Detecting and Resolving Code-Comment Inconsistencies (2024.eacl-demo)
Copied to clipboard
| Challenge: | Existing methods to detect and rectify inconsistencies in source code rely on heuristic rules . however, there are limitations to generating synthetic comments . |
| Approach: | They introduce a language model-based framework capable of detecting inconsistencies between code and comments. |
| Outcome: | The proposed framework achieves 72.3% accuracy and scores 33.64 on the code summarization task. |