The field of software development is rapidly evolving with the integration of Artificial Intelligence (AI) and Large Language Models (LLMs). Recent studies have demonstrated the potential of LLMs in assisting software development tasks, such as code generation, bug fixing, and code explanation. The use of LLMs has shown promise in improving the efficiency and accuracy of these tasks. Notably, the development of frameworks like P4OMP, RAILS, and QLPro has enabled the effective integration of LLMs with software development tools, resulting in improved code generation and bug fixing capabilities. Furthermore, the application of LLMs in code explanation and code search has also shown significant improvements. Overall, the field is moving towards the development of more advanced and specialized AI-assisted software development tools. Noteworthy papers in this area include P4OMP, which achieves 100% compilation success in generating OpenMP-annotated parallel code, and RAILS, which outperforms baseline prompting in preserving intent and avoiding hallucinations. QLPro is also notable for its ability to detect 41 confirmed vulnerabilities in open-source projects, including 6 previously unknown vulnerabilities.
Advances in AI-Assisted Software Development
Sources
An LLM-assisted approach to designing software architectures using ADD
P4OMP: Retrieval-Augmented Prompting for OpenMP Parallelism in Serial Code
RAILS: Retrieval-Augmented Intelligence for Learning Software Development
Privacy-Preserving Methods for Bug Severity Prediction
Evaluating and Improving Large Language Models for Competitive Program Generation
Guiding AI to Fix Its Own Flaws: An Empirical Study on LLM-Driven Secure Code Generation
Repair Ingredients Are All You Need: Improving Large Language Model-Based Program Repair via Repair Ingredients Search
On the Feasibility of Deduplicating Compiler Bugs with Bisection
Compiling a Q# Subset to QASM 3.0 in TypeScript via a JSON Based IR
Comparative Analysis of the Code Generated by Popular Large Language Models (LLMs) for MISRA C++ Compliance
QLPro: Automated Code Vulnerability Discovery via LLM and Static Code Analysis Integration
A Survey of LLM-based Automated Program Repair: Taxonomies, Design Paradigms, and Applications
Bug Fixing with Broader Context: Enhancing LLM-Based Program Repair via Layered Knowledge Injection
Recommending Variable Names for Extract Local Variable Refactorings
A Hierarchical and Evolvable Benchmark for Fine-Grained Code Instruction Following with Multi-Turn Feedback
Context-Aware Code Wiring Recommendation with LLM-based Agent
DaiFu: In-Situ Crash Recovery for Deep Learning Systems
APRMCTS: Improving LLM-based Automated Program Repair with Iterative Tree Search
Structural Code Search using Natural Language Queries
Enhancing COBOL Code Explanations: A Multi-Agents Approach Using Large Language Models