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