CodeJudge-Eval: Can Large Language Models be Good Judges in Code Understanding?

TL;DR

Eval Evaluation of 12 well-known LLMs on CJ-Eval reveals that even state-of-the-art models struggle, highlighting the benchmark's ability to probe deeper into models' code understanding abilities.

Abstract

Recent advancements in large language models (LLMs) have showcased impressive code generation capabilities, primarily evaluated through language-to-code benchmarks. However, these benchmarks may not fully capture a model's code understanding abilities. We introduce CodeJudge-Eval (CJ-Eval), a novel benchmark designed to assess LLMs' code understanding abilities from the perspective of code judging rather than code generation. CJ-Eval challenges models to determine the correctness of provided code solutions, encompassing various error types and compilation issues. By leveraging a diverse set of problems and a fine-grained judging system, CJ-Eval addresses the limitations of traditional benchmarks, including the potential memorization of solutions. Evaluation of 12 well-known LLMs on CJ-Eval reveals that even state-of-the-art models struggle, highlighting the benchmark's ability to probe deeper into models' code understanding abilities. Our codes and benchmark are available at \url{https://github.com/CodeLLM-Research/CodeJudge-Eval}.

Figures (13)

• Figure 1: Comparing code generation with code judging task, we observe that a model's ability to generate correct code does not necessarily imply it can accurately judge other codes for the same problem.
• Figure 2: An overview of our pipeline for constructing the CodeJudge-Eval benchmark.
• Figure 3: A stacked histogram on the number of test cases in the filtered problems. Different filtering thresholds are applied based on different difficulty.
• Figure 4: Scaling the number of models' parameters on our CJ-Eval Easy.
• Figure 5: Comparative few-shot F1 scores of different models on our CJ-Eval easy.