Code-Centric Detection of Vulnerability-Fixing Commits: A Unified Benchmark and Empirical Study

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Loose, Nils, Bienhüls, Joseph, Hempel, Kristoffer, Mächtle, Felix, Eisenbarth, Thomas
Natura: Preprint
Pubblicazione: 2026
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866917490741739520
author Loose, Nils
Bienhüls, Joseph
Hempel, Kristoffer
Mächtle, Felix
Eisenbarth, Thomas
author_facet Loose, Nils
Bienhüls, Joseph
Hempel, Kristoffer
Mächtle, Felix
Eisenbarth, Thomas
contents Automated detection of vulnerability-fixing commits (VFCs) is critical for timely security patch deployment, as advisory databases lag patch releases by a median of 25 days and many fixes never receive advisories. We present a comprehensive evaluation of code language model based VFC detection through a unified framework consolidating over 20 fragmented datasets spanning more than 180000 commits. Across over 180 experiments with fine-tuned models from 125 M to 14 B parameters, we find no evidence that models acquire transferable security-relevant code understanding from code changes alone. When commit messages are available, they dominate model attention, and when removed, an attribution analysis shows that enriching diffs with additional intra-procedural semantic context does not shift model attention toward the code changes. Group-stratified evaluation exposes approximately 17% performance drops compared to random splits, while temporal splits on aggregated datasets prove unreliable due to compositional shift in the underlying project distributions. At a false positive rate of 0.5% all fine-tuned code-only models miss over 93% of vulnerabilities. Larger and more diverse training data or generative approaches show preliminary improvements but do not resolve the underlying limitations. To support future research on code-centric VFC detection, we release our unified framework and evaluation suite.
format Preprint
id arxiv_https___arxiv_org_abs_2605_13138
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Code-Centric Detection of Vulnerability-Fixing Commits: A Unified Benchmark and Empirical Study
Loose, Nils
Bienhüls, Joseph
Hempel, Kristoffer
Mächtle, Felix
Eisenbarth, Thomas
Software Engineering
Cryptography and Security
Machine Learning
Automated detection of vulnerability-fixing commits (VFCs) is critical for timely security patch deployment, as advisory databases lag patch releases by a median of 25 days and many fixes never receive advisories. We present a comprehensive evaluation of code language model based VFC detection through a unified framework consolidating over 20 fragmented datasets spanning more than 180000 commits. Across over 180 experiments with fine-tuned models from 125 M to 14 B parameters, we find no evidence that models acquire transferable security-relevant code understanding from code changes alone. When commit messages are available, they dominate model attention, and when removed, an attribution analysis shows that enriching diffs with additional intra-procedural semantic context does not shift model attention toward the code changes. Group-stratified evaluation exposes approximately 17% performance drops compared to random splits, while temporal splits on aggregated datasets prove unreliable due to compositional shift in the underlying project distributions. At a false positive rate of 0.5% all fine-tuned code-only models miss over 93% of vulnerabilities. Larger and more diverse training data or generative approaches show preliminary improvements but do not resolve the underlying limitations. To support future research on code-centric VFC detection, we release our unified framework and evaluation suite.
title Code-Centric Detection of Vulnerability-Fixing Commits: A Unified Benchmark and Empirical Study
topic Software Engineering
Cryptography and Security
Machine Learning
url https://arxiv.org/abs/2605.13138