SOURCE-LINKED INTELLIGENCE

Zeva-Ego: Egocentric Mid-Training with In-Context Causal Learning for Robot Manipulation

arXiv · AI, language, vision and robotics · article · Sep 21, 2026 · UTC

Egocentric video offers a scalable source of physical interaction experience, yet translating it into robot-executable knowledge and enabling continual adaptation remain challenging. We introduce Zeva-Ego, a unified framework that learns physical priors from human experience and evolves through robot interaction. An Action-Centric Encoder (ACE) converts egocentric visual transitions into action-centered supervision for VLA mid-training, while In-Context Causal Learning (ICCL) enables parameter-free adaptation from action-effect feedback at deployment. Scaling Ego data to 10K hours improves Rob

Read original source ↗ Open in workspace

recordType

paper

region

Global