Decompile-Bench: Million-Scale Binary-Source Function Pairs for Real-World Binary Decompilation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Tan, Hanzhuo, Tian, Xiaolong, Qi, Hanrui, Liu, Jiaming, Gao, Zuchen, Wang, Siyi, Luo, Qi, Li, Jing, Zhang, Yuqun
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918163492372480
author Tan, Hanzhuo
Tian, Xiaolong
Qi, Hanrui
Liu, Jiaming
Gao, Zuchen
Wang, Siyi
Luo, Qi
Li, Jing
Zhang, Yuqun
author_facet Tan, Hanzhuo
Tian, Xiaolong
Qi, Hanrui
Liu, Jiaming
Gao, Zuchen
Wang, Siyi
Luo, Qi
Li, Jing
Zhang, Yuqun
contents Recent advances in LLM-based decompilers have been shown effective to convert low-level binaries into human-readable source code. However, there still lacks a comprehensive benchmark that provides large-scale binary-source function pairs, which is critical for advancing the LLM decompilation technology. Creating accurate binary-source mappings incurs severe issues caused by complex compilation settings and widespread function inlining that obscure the correspondence between binaries and their original source code. Previous efforts have either relied on used contest-style benchmarks, synthetic binary-source mappings that diverge significantly from the mappings in real world, or partially matched binaries with only code lines or variable names, compromising the effectiveness of analyzing the binary functionality. To alleviate these issues, we introduce Decompile-Bench, the first open-source dataset comprising two million binary-source function pairs condensed from 100 million collected function pairs, i.e., 450GB of binaries compiled from permissively licensed GitHub projects. For the evaluation purposes, we also developed a benchmark Decompile-Bench-Eval including manually crafted binaries from the well-established HumanEval and MBPP, alongside the compiled GitHub repositories released after 2025 to mitigate data leakage issues. We further explore commonly-used evaluation metrics to provide a thorough assessment of the studied LLM decompilers and find that fine-tuning with Decompile-Bench causes a 20% improvement over previous benchmarks in terms of the re-executability rate. Our code and data has been released in HuggingFace and Github. https://github.com/albertan017/LLM4Decompile
format Preprint
id arxiv_https___arxiv_org_abs_2505_12668
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Decompile-Bench: Million-Scale Binary-Source Function Pairs for Real-World Binary Decompilation
Tan, Hanzhuo
Tian, Xiaolong
Qi, Hanrui
Liu, Jiaming
Gao, Zuchen
Wang, Siyi
Luo, Qi
Li, Jing
Zhang, Yuqun
Software Engineering
Recent advances in LLM-based decompilers have been shown effective to convert low-level binaries into human-readable source code. However, there still lacks a comprehensive benchmark that provides large-scale binary-source function pairs, which is critical for advancing the LLM decompilation technology. Creating accurate binary-source mappings incurs severe issues caused by complex compilation settings and widespread function inlining that obscure the correspondence between binaries and their original source code. Previous efforts have either relied on used contest-style benchmarks, synthetic binary-source mappings that diverge significantly from the mappings in real world, or partially matched binaries with only code lines or variable names, compromising the effectiveness of analyzing the binary functionality. To alleviate these issues, we introduce Decompile-Bench, the first open-source dataset comprising two million binary-source function pairs condensed from 100 million collected function pairs, i.e., 450GB of binaries compiled from permissively licensed GitHub projects. For the evaluation purposes, we also developed a benchmark Decompile-Bench-Eval including manually crafted binaries from the well-established HumanEval and MBPP, alongside the compiled GitHub repositories released after 2025 to mitigate data leakage issues. We further explore commonly-used evaluation metrics to provide a thorough assessment of the studied LLM decompilers and find that fine-tuning with Decompile-Bench causes a 20% improvement over previous benchmarks in terms of the re-executability rate. Our code and data has been released in HuggingFace and Github. https://github.com/albertan017/LLM4Decompile
title Decompile-Bench: Million-Scale Binary-Source Function Pairs for Real-World Binary Decompilation
topic Software Engineering
url https://arxiv.org/abs/2505.12668