Abstract: Deep learning (DL) techniques have demonstrated potential in reasoning complex patterns of vulnerable code from high-level abstractions. Recent advancements in the area, such as the ...