Transformer Transformer: A Unified Model for Motion-Conditioned Robot Co-design
Diffusion transformer over a shared tokenization of robot bodies, states and actions that generates robot designs to track target motions. A fabricated optimised ALOHA design cut tracking error by over 70%.基于统一的机器人本体/状态/动作 token 表示的扩散 Transformer，按目标末端轨迹生成机器人设计。实际制作的优化版 ALOHA 跟踪误差降低 70% 以上。
BibTeX
@misc{ha2026transformer,
title = {{Transformer Transformer: A Unified Model for Motion-Conditioned Robot Co-design}},
author = {Huy Ha and C. Karen Liu and Shuran Song},
year = {2026},
eprint = {2607.25798},
archivePrefix = {arXiv},
primaryClass = {cs.RO},
url = {https://arxiv.org/abs/2607.25798},
}Hardware & Open-Source Robots硬件与开源机器人co-designdiffusion-transformerembodiment-generationalohafabrication