Beyond Output Correctness: Benchmarking and Evaluating Large Language Model Reasoning in Coding Tasks

TL;DR

This paper introduces CodeRQ-Bench, a new benchmark for evaluating reasoning in LLMs across coding tasks, and proposes VERA, a two-stage evaluator that improves reasoning assessment accuracy.

Contribution

It presents the first comprehensive benchmark for reasoning in coding tasks and introduces VERA, a novel evaluation method guided by insights from analyzing existing evaluators.

Findings

VERA outperforms strong baselines on multiple datasets.

CodeRQ-Bench is publicly available for future research.

Analysis reveals five limitations in current reasoning evaluators.

Abstract

Large language models (LLMs) increasingly rely on explicit reasoning to solve coding tasks, yet evaluating the quality of this reasoning remains challenging. Existing reasoning evaluators are not designed for coding, and current benchmarks focus primarily on code generation, leaving other coding tasks largely unexplored. We introduce CodeRQ-Bench, the first benchmark for evaluating LLM reasoning quality across three coding task categories: generation, summarization, and classification. Using this benchmark, we analyze 1,069 mismatch cases from existing evaluators, identify five recurring limitations, and derive four design insights for reasoning evaluation in coding tasks. Guided by these insights, we propose VERA, a two-stage evaluator that combines evidence-grounded verification with ambiguity-aware score correction. Experiments on CodeRQ-Bench show that VERA consistently outperforms strong baselines across four datasets, improving AUCROC by up to 0.26 and AUPRC by up to 0.21. We release CodeRQ-Bench at https://github.com/MrLYG/CodeRQ-Bench, supporting future investigations.