The field of software development is witnessing significant advancements with the integration of Large Language Models (LLMs). Recent studies have explored the application of LLMs in various aspects of software development, including code review, code translation, and code generation. A notable trend is the focus on improving the efficiency and effectiveness of LLMs in these tasks, with techniques such as fine-tuning and prompting being employed to enhance their performance. Additionally, there is a growing emphasis on evaluating the quality and security of LLM-generated code, with metrics such as correctness, efficiency, and maintainability being used to assess their capabilities. Noteworthy papers in this area include TRACY, which introduces a comprehensive benchmark for evaluating the execution efficiency of LLM-translated code, and COMPASS, which proposes a multi-dimensional evaluation framework for assessing code generation in LLMs. Overall, the field is moving towards developing more robust and reliable LLMs that can be effectively utilized in software development tasks.
Advancements in Large Language Models for Software Development
Sources
The Impact of Large Language Models (LLMs) on Code Review Process
TRACY: Benchmarking Execution Efficiency of LLM-Based Code Translation
Large Language Models in the Data Science Lifecycle: A Systematic Mapping Study
Code Vulnerability Detection Across Different Programming Languages with AI Models
WIP: Leveraging LLMs for Enforcing Design Principles in Student Code: Analysis of Prompting Strategies and RAG
Clean Code, Better Models: Enhancing LLM Performance with Smell-Cleaned Dataset
How Much Can a Behavior-Preserving Changeset Be Decomposed into Refactoring Operations?
LinkAnchor: An Autonomous LLM-Based Agent for Issue-to-Commit Link Recovery
XAMT: Cross-Framework API Matching for Testing Deep Learning Libraries
Strengthening Programming Comprehension in Large Language Models through Code Generation
ChangePrism: Visualizing the Essence of Code Changes
RUM: Rule+LLM-Based Comprehensive Assessment on Testing Skills
The Hidden Cost of Readability: How Code Formatting Silently Consumes Your LLM Budget
COMPASS: A Multi-Dimensional Benchmark for Evaluating Code Generation in Large Language Models
Prompt Orchestration Markup Language
Measuring LLM Code Generation Stability via Structural Entropy
Static Analysis as a Feedback Loop: Enhancing LLM-Generated Code Beyond Correctness
Assessing the Quality and Security of AI-Generated Code: A Quantitative Analysis
On the need to perform comprehensive evaluations of automated program repair benchmarks: Sorald case study
LLaSO: A Foundational Framework for Reproducible Research in Large Language and Speech Model
SynthCoder: A Synthetical Strategy to Tune LLMs for Code Completion
Evaluation Guidelines for Empirical Studies in Software Engineering involving LLMs
Towards Scalable and Interpretable Mobile App Risk Analysis via Large Language Models