arXiv · 2611.21626
X-LocVul: An Evidence-Propagation Audit of Cascaded C/C++ Vulnerability Detection, Localization, and Explanation
Abstract
Developer-facing vulnerability analysis should link a function-level warning to a source location and a supported explanation. Cascaded systems can propagate weak evidence despite plausible components. We present X-LocVul, a measurement-first audit of a modular C/C++ detector-localizer-explainer pipeline combining UniXCoder, CodeT5, and Qwen2.5-Coder. We test whether evidence passed across stages is useful, merely source-resident, or misleading. Under duplication and temporal-leakage controls, UniXCoder has higher point estimates than CodeBERT, but true-CWE supervision is statistically indistinguishable from a shuffled-CWE control on PR-AUC, MCC, and vulnerable-class F1. Raw generated statements fail exact source matching in 90.83% of cases. Fuzzy projection reaches 83.79% Top-1 accuracy and exceeds semantic projection by 1.59 percentage points, although its advantage over raw ranking is not established. In a randomized, condition-label-masked two-rater evaluation of 50 paired cases, a predicted line improves evidence support, whereas a predicted CWE improves no endpoint. Oracle evidence improves evidence support (Holm-adjusted p=0.00453) and CWE consistency (p=3.49 x 10^-7), but not root-cause quality after correction (p=0.250); unsupported claims remain common. A chronological stress test retains ranking signal (ROC-AUC 0.782; PR-AUC 0.110), while the transferred threshold yields zero recall and F1. The contribution is not a state-of-the-art claim, but an auditable account of evidence construction, transfer, and consumption. It separates source residency from localization correctness, auxiliary labels from semantic supervision, and predicted evidence from useful evidence. Public E4 scores regenerate all 15 tests; missing manifests, final prompts, and profiler logs bound stronger reproducibility claims.
Explore related subjects
Keep this discovery
Explore connections, maps & timelines
Hoang-Phuong-Linh Le, Quyet-Thang Huynh. 2026-09-13. X-LocVul: An Evidence-Propagation Audit of Cascaded C/C++ Vulnerability Detection, Localization, and Explanation. https://arxiv.science/abs/2611.21626
Citations retain the original authors and DOI and include this site’s record URL. If this discovery service contributed to your research, you can also cite arXiv Science by Indiciae: BibTeX · RIS.