First-author paper, “Revisiting Pre-trained Language Models for Vulnerability Detection”, accepted to the ACM ASIA Conference on Computer and Communications Security (ASIACCS ‘26).