Temporal Vision Transformers for Generalizable Detection and Localization of AI-Generated Video Deepfakes in the Wild. AMAR. 2026;4(3):1445-1466. doi:10.5281/zenodo.23189092