CodeIF-Bench: Evaluating Instruction-Following Capabilities of Large Language Models in Interactive Code Generation

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Wang, Peiding, Zhang, Li, Liu, Fang, Shi, Lin, Li, Minxiao, Shen, Bo, Fu, An
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866917099090214912
author Wang, Peiding
Zhang, Li
Liu, Fang
Shi, Lin
Li, Minxiao
Shen, Bo
Fu, An
author_facet Wang, Peiding
Zhang, Li
Liu, Fang
Shi, Lin
Li, Minxiao
Shen, Bo
Fu, An
contents Large Language Models (LLMs) have demonstrated exceptional performance in code generation tasks and have become indispensable programming assistants for developers. However, existing code generation benchmarks primarily assess the functional correctness of code generated by LLMs in single-turn interactions. They offer limited insight into LLMs' abilities to generate code that strictly follows users' instructions in multi-turn interaction scenarios. In this paper, we introduce CodeIF-Bench, a benchmark for evaluating the instruction-following capabilities of LLMs in interactive code generation. Specifically, CodeIF-Bench incorporates nine types of verifiable instructions aligned with the real-world software development requirements, which can be independently and objectively validated through specified test cases, facilitating the evaluation of instruction-following capability in multi-turn interactions. In both Static Conversation and Dynamic Conversation settings, we evaluate the performance of 6 state-of-the-art LLMs and summarize the important factors, additional repository context and gradually increasing interaction history influencing the instruction-following ability of LLMs in multi-turn interactions. Furthermore, we identify the potential direction for improvement: context management. The code and data are available at \href{https://github.com/zhu-zhu-ding/CodeIF-Bench}{https://github.com/zhu-zhu-ding/CodeIF-Bench}.
format Preprint
id arxiv_https___arxiv_org_abs_2503_22688
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle CodeIF-Bench: Evaluating Instruction-Following Capabilities of Large Language Models in Interactive Code Generation
Wang, Peiding
Zhang, Li
Liu, Fang
Shi, Lin
Li, Minxiao
Shen, Bo
Fu, An
Software Engineering
Artificial Intelligence
Programming Languages
Large Language Models (LLMs) have demonstrated exceptional performance in code generation tasks and have become indispensable programming assistants for developers. However, existing code generation benchmarks primarily assess the functional correctness of code generated by LLMs in single-turn interactions. They offer limited insight into LLMs' abilities to generate code that strictly follows users' instructions in multi-turn interaction scenarios. In this paper, we introduce CodeIF-Bench, a benchmark for evaluating the instruction-following capabilities of LLMs in interactive code generation. Specifically, CodeIF-Bench incorporates nine types of verifiable instructions aligned with the real-world software development requirements, which can be independently and objectively validated through specified test cases, facilitating the evaluation of instruction-following capability in multi-turn interactions. In both Static Conversation and Dynamic Conversation settings, we evaluate the performance of 6 state-of-the-art LLMs and summarize the important factors, additional repository context and gradually increasing interaction history influencing the instruction-following ability of LLMs in multi-turn interactions. Furthermore, we identify the potential direction for improvement: context management. The code and data are available at \href{https://github.com/zhu-zhu-ding/CodeIF-Bench}{https://github.com/zhu-zhu-ding/CodeIF-Bench}.
title CodeIF-Bench: Evaluating Instruction-Following Capabilities of Large Language Models in Interactive Code Generation
topic Software Engineering
Artificial Intelligence
Programming Languages
url https://arxiv.org/abs/2503.22688