---
task_categories:
- robotics
tags:
- rlbench
- colosseum
- gembench
---

[BridgeVLA: Input-Output Alignment for Efficient 3D Manipulation Learning with Vision-Language Models](https://huggingface.co/papers/2506.07961)

Project Website: https://bridgevla.github.io/

arxiv: https://arxiv.org/abs/2506.07961