# RynnVLA-Latent: Scaling Robotic Manipulation Foundation Models with 3D-Grounded Latent Actions

- Source: arXiv cs.RO
- Published: 2026-10-27T04:00:00.000Z
- RoboHot: https://robohot.news/en/items/egqqg5w1huf9edm8qvr1annuh
- Original: https://arxiv.org/abs/2610.34356

## Summary

The research team proposes the RynnVLA-Latent framework, which uses RynnLAM to learn continuous, 3D-grounded latent actions from visual changes without native robot control signals, converting 30.1K hours of human, robot, and simulation video into 745.52M latent action descriptors for policy pretraining.

## Why it matters

This research uses 3D-grounded latent actions to turn heterogeneous interaction videos into pretraining supervision, offering a reusable path for VLA models to bypass the scarcity of action annotations.