SOURCE-LINKED INTELLIGENCE

Mask-Aware Execution for Efficient JEPA Training

arXiv · AI, language, vision and robotics · article · Sep 19, 2026 · UTC

Joint Embedding Predictive Architectures (JEPAs) are becoming a core representation-learning primitive and a building block for latent world models across vision, video, audio, brain dynamics, and time series. Despite (potential of) wide deployment, current JEPA training pipelines are inefficient: each input is executed through multiple mask-specific branches, with redundant target-side work, and memory-bound token routing. These costs grow with the number of masks and limit GPU efficiency. We present M-JEPA, a mask-aware execution architecture that restructures JEPA training without changing

Read original source ↗ Open in workspace

recordType

paper

region

Global