Can LLMs Generate High-Quality Test Cases for Algorithm Problems? TestCase-Eval: A Systematic Evaluation of Fault Coverage and Exposure

TL;DR

This paper introduces TestCase-Eval, a comprehensive benchmark for evaluating how effectively large language models generate test cases for algorithm problems, focusing on fault coverage and exposure.

Contribution

It presents a new benchmark with a systematic evaluation framework for assessing LLMs' ability to generate high-quality test cases for algorithm problems.

Findings

LLMs vary significantly in fault coverage and exposure capabilities.

Proprietary LLMs outperform open-source models in test case generation.

The benchmark reveals specific strengths and limitations of current LLMs.

Abstract

We introduce TestCase-Eval, a new benchmark for systematic evaluation of LLMs in test-case generation. TestCase-Eval includes 500 algorithm problems and 100,000 human-crafted solutions from the Codeforces platform. It focuses on two pivotal tasks: (1) Fault Coverage, which measures how well LLM-generated test sets probe diverse input scenarios and cover a wide range of potential failure modes. (2) Fault Exposure, which evaluates whether LLMs can craft a tailored test input that reveals a specific incorrect code implementation. We provide a comprehensive assessment of 19 state-of-the-art open-source and proprietary LLMs on TestCase-Eval, offering insights into their strengths and limitations in generating effective test cases for algorithm problems.