Author

Mohan Liu

View on Semantic Scholar

1 paper indexed here

We haven’t gathered this author’s papers yet. Follow them and we’ll fetch their work.

Not the right person? Other researchers publish under this name.

Jul 2026

VistaVLA: Geometry- and Semantic-Aware 3D Gaussian-Grounded VLA for Robotic Manipulation

This work presents VistaVLA, a novel two-stage framework that constructs a geometry- and semantics-aware 3D cognitive representation from 3D Gaussian primitives and grounds it as compact context tokens for VLA policy learning, and introduces Merge-then-Query (MtQ), a token summarization mechanism.

Mo-Han Liu, Zhihao Gu, Xuanyu Chen et al. · arXiv.org · 0 citations