Benchmarks and Metrics for Evaluations of Code Generation: A Critical Review

TL;DR

This paper critically reviews existing benchmarks and metrics used to evaluate large language models for code generation, highlighting gaps and proposing future research directions.

Contribution

It provides a comprehensive analysis of current evaluation methods for code generation LLMs and discusses potential improvements and research directions.

Findings

Current benchmarks lack standardization.

Metrics often do not fully capture code quality.

Future research should focus on more comprehensive evaluation methods.

Abstract

With the rapid development of Large Language Models (LLMs), a large number of machine learning models have been developed to assist programming tasks including the generation of program code from natural language input. However, how to evaluate such LLMs for this task is still an open problem despite of the great amount of research efforts that have been made and reported to evaluate and compare them. This paper provides a critical review of the existing work on the testing and evaluation of these tools with a focus on two key aspects: the benchmarks and the metrics used in the evaluations. Based on the review, further research directions are discussed.