TokenGated-CodeBERT With Lightweight Attention-Based Token Selection for Software Defect Prediction

Class imbalance poses a key challenge in defective code module detection, severely limiting the performance of deep learning models. To address this issue, we propose TokenGated-CodeBERT, a new methodology designed for scenarios with severe class imbalance. Constructed on the basis of vanilla CodeBE...

Descripción completa

Detalles Bibliográficos

Autores principales:	Bin Shuai, Dequan Xu, Yuanlin Yang
Formato:	Artigo
Lenguaje:	Inglês
Publicado:	IEEE 2026-01-01
Colección:	IEEE Access
Materias:	Software defect prediction CodeBERT software reliability deep learning
Acceso en línea:	https://ieeexplore.ieee.org/document/11594087/
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!