LLM Hallucinations in Practical Code Generation: Phenomena, Mechanism, and Mitigation
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2024
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866909458340249600 |
|---|---|
| author | Zhang, Ziyao Wang, Yanlin Wang, Chong Chen, Jiachi Zheng, Zibin |
| author_facet | Zhang, Ziyao Wang, Yanlin Wang, Chong Chen, Jiachi Zheng, Zibin |
| contents | Code generation aims to automatically generate code from input requirements, significantly enhancing development efficiency. Recent large language models (LLMs) based approaches have shown promising results and revolutionized code generation task. Despite the promising performance, LLMs often generate contents with hallucinations, especially for the code generation scenario requiring the handling of complex contextual dependencies in practical development process. Although previous study has analyzed hallucinations in LLM-powered code generation, the study is limited to standalone function generation. In this paper, we conduct an empirical study to study the phenomena, mechanism, and mitigation of LLM hallucinations within more practical and complex development contexts in repository-level generation scenario. First, we manually examine the code generation results from six mainstream LLMs to establish a hallucination taxonomy of LLM-generated code. Next, we elaborate on the phenomenon of hallucinations, analyze their distribution across different models. We then analyze causes of hallucinations and identify four potential factors contributing to hallucinations. Finally, we propose an RAG-based mitigation method, which demonstrates consistent effectiveness in all studied LLMs. The replication package including code, data, and experimental results is available at https://github.com/DeepSoftwareAnalytics/LLMCodingHallucination |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2409_20550 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | LLM Hallucinations in Practical Code Generation: Phenomena, Mechanism, and Mitigation Zhang, Ziyao Wang, Yanlin Wang, Chong Chen, Jiachi Zheng, Zibin Software Engineering Artificial Intelligence Computation and Language Code generation aims to automatically generate code from input requirements, significantly enhancing development efficiency. Recent large language models (LLMs) based approaches have shown promising results and revolutionized code generation task. Despite the promising performance, LLMs often generate contents with hallucinations, especially for the code generation scenario requiring the handling of complex contextual dependencies in practical development process. Although previous study has analyzed hallucinations in LLM-powered code generation, the study is limited to standalone function generation. In this paper, we conduct an empirical study to study the phenomena, mechanism, and mitigation of LLM hallucinations within more practical and complex development contexts in repository-level generation scenario. First, we manually examine the code generation results from six mainstream LLMs to establish a hallucination taxonomy of LLM-generated code. Next, we elaborate on the phenomenon of hallucinations, analyze their distribution across different models. We then analyze causes of hallucinations and identify four potential factors contributing to hallucinations. Finally, we propose an RAG-based mitigation method, which demonstrates consistent effectiveness in all studied LLMs. The replication package including code, data, and experimental results is available at https://github.com/DeepSoftwareAnalytics/LLMCodingHallucination |
| title | LLM Hallucinations in Practical Code Generation: Phenomena, Mechanism, and Mitigation |
| topic | Software Engineering Artificial Intelligence Computation and Language |
| url | https://arxiv.org/abs/2409.20550 |