Towards Reliable Machine Learning Models for Code

• Item type: Symposium(1)
• 公開日: 2024-09-10
• タイトル: Towards Reliable Machine Learning Models for Code
• タイトル / 言語: en
• タイトル: Towards Reliable Machine Learning Models for Code
• 言語: eng
• キーワード / 主題Scheme: Other
• キーワード / 主題: 国際セッション
• 資源タイプ / 資源タイプ識別子: http://purl.org/coar/resource_type/c_5794
• 資源タイプ: conference paper
• 著者所属: Polytechnique Montréal
• 著者所属(英): en
• 著者所属(英): Polytechnique Montréal
• 著者名: Foutse, Khomh
• 著者名(英): Foutse, Khomh
• 論文抄録 / 内容記述タイプ: Other
• 論文抄録 / 内容記述: Machine learning (ML) models trained on code are increasingly integrated into various software engineering tasks. While they generally demonstrate promising performance, many aspects of their capabilities remain unclear. Specifically, there is a lack of understanding regarding what these models learn, why they learn it, how they operate, and when they produce erroneous outputs. In this talk, I will present findings from a series of studies that (i) examine the abilities of these models to complement human developers, (ii) explore the syntax and representation learning capabilities of ML models designed for software maintenance tasks, and (iii) investigate the patterns of bugs these models exhibit. Additionally, I will discuss a novel self-refinement approach aimed at enhancing the reliability of code generated by Large Language Models (LLMs). This method focuses on reducing the occurrence of bugs before execution, autonomously and without the need for human intervention or predefined test cases.
• 論文抄録(英) / 内容記述タイプ: Other
• 論文抄録(英) / 内容記述: Machine learning (ML) models trained on code are increasingly integrated into various software engineering tasks. While they generally demonstrate promising performance, many aspects of their capabilities remain unclear. Specifically, there is a lack of understanding regarding what these models learn, why they learn it, how they operate, and when they produce erroneous outputs. In this talk, I will present findings from a series of studies that (i) examine the abilities of these models to complement human developers, (ii) explore the syntax and representation learning capabilities of ML models designed for software maintenance tasks, and (iii) investigate the patterns of bugs these models exhibit. Additionally, I will discuss a novel self-refinement approach aimed at enhancing the reliability of code generated by Large Language Models (LLMs). This method focuses on reducing the occurrence of bugs before execution, autonomously and without the need for human intervention or predefined test cases.
• 書誌情報: ソフトウェアエンジニアリングシンポジウム2024論文集; 巻 2024, p. 17-17, 発行日 2024-09-10
• 出版者 / 言語: ja
• 出版者: 情報処理学会