A Qualitative Investigation into LLM-Generated Multilingual Code Comments and Automatic Evaluation Metrics
This work addresses the problem of unreliable automatic evaluation metrics for multilingual code comments, which is incremental as it builds on existing models and metrics.
The study evaluated five state-of-the-art code language models on generating code comments in five languages, identifying 26 error categories and showing that modern neural metrics fail to reliably distinguish correct from incorrect comments, with significant score overlap between expert-rated categories.
Large Language Models are essential coding assistants, yet their training is predominantly English-centric. In this study, we evaluate the performance of code language models in non-English contexts, identifying challenges in their adoption and integration into multilingual workflows. We conduct an open-coding study to analyze errors in code comments generated by five state-of-the-art code models, CodeGemma, CodeLlama, CodeQwen1.5, GraniteCode, and StarCoder2 across five natural languages: Chinese, Dutch, English, Greek, and Polish. Our study yields a dataset of 12,500 labeled generations, which we publicly release. We then assess the reliability of standard metrics in capturing comment \textit{correctness} across languages and evaluate their trustworthiness as judgment criteria. Through our open-coding investigation, we identified a taxonomy of 26 distinct error categories in model-generated code comments. They highlight variations in language cohesion, informativeness, and syntax adherence across different natural languages. Our analysis shows that, while these models frequently produce partially correct comments, modern neural metrics fail to reliably differentiate meaningful completions from random noise. Notably, the significant score overlap between expert-rated correct and incorrect comments calls into question the effectiveness of these metrics in assessing generated comments.