Zheng Xiong, Kang Li, Zilin Wang, Matthew Jackson, Jakob Foerster, Shimon Whiteson
HyperVLA: Efficient Inference in Vision-Language-Action Models via Hypernetworks
https://arxiv.org/abs/2510.04898

HyperVLA reduces inference load by activating only about 1% of VLA model parameters, delivering roughly a 120× speed boost while keeping zero‑shot success rates comparable. Read more: https://getnews.me/hypervla-cuts-vision-language-action-model-inference-cost-by-90x/ #hypervla #visionlanguageaction

Zheng Xiong, Kang Li, Zilin Wang, Matthew Jackson, Jakob Foerster, Shimon Whiteson: HyperVLA: Efficient Inference in Vision-Language-Action Models via Hypernetworks https://arxiv.org/abs/2510.04898 https://arxiv.org/pdf/2510.04898 https://arxiv.org/html/2510.04898