Combining TF-IDF token n-grams with code metrics like lines of code and complexity in a logistic regression model yields PR-AUC of 0.642 on random splits of Devign but lower performance in cross-project settings.
VulDeePecker: A Deep Learning-Based System for Vulnerability Detection
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CR 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Lightweight Vulnerability Detection from Code Metrics and Token Features
Combining TF-IDF token n-grams with code metrics like lines of code and complexity in a logistic regression model yields PR-AUC of 0.642 on random splits of Devign but lower performance in cross-project settings.