ALT The Image-based Joint-Embedding Predictive Architecture (I-JEPA) uses a single context block to predict the representations of various target blocks originating from the same image. The context encoder is a Vision Transformer (ViT) that only processes the visible context patches. The predictor is a narrow ViT that takes the context encoder output and predicts the representations of a target block at a specific location, conditioned on positional tokens of the target (shown in color). The target representations correspond to the outputs of the target-encoder, the weights of which are updated at each iteration via an exponential moving average of the context encoder weights.