Latent prediction error does not always indicate physical surprise
Framework separates physical surprise from appearance and motion in video models
20
Framework separates physical surprise from appearance and motion in video models
TT-VidT pairs a DINOv3 spatial ViT with a temporal layer, leading motion benchmarks at lower FLOPs.
Today 1.2K readers · 1.4K views · 189 reads · all time 203K views