Meta FAIR researchers posted RoboJEPA, an 8-billion-parameter Joint Embedding Predictive Architecture world model trained on real-robot data across 12 embodiments. The paper, dated Oct. 7, 2026 and submitted to Hugging Face on Oct. 8, says imagination error follows a compute scaling law and that checkpoints plus training and robot-deployment code are released. Correspondence lists Jeannette Bohg at Meta; the release is a research artifact, not a Meta consumer-product launch.
/ Sources
/ About this story
Compiled by Venture Atlas from the sources above, using automated AI-assisted research. This is a summary of reporting published elsewhere, not original reporting - follow the source links for the full account. See our editorial standards.
Something wrong here? Email apostructura@gmail.com and we will fix it.