HMVLA: Hyperbolic Multimodal Fusion for Vision-Language-Action Models

作者：Kun Wang, Xiao Feng, Mingcheng Qu, Tonghua Su · 单位：Harbin Institute of Technology, Harbin, China, Chongqing Research Institute of HIT, Chongqing, China, Guangdong Laboratory of Artificial Intelligence and Digital Economy (SZ), Shenzhen, China · 会议/期刊：ICASSP 2026 · 日期：2026-01-28 · 来源：Low-Level Learning-Based Action Modelling / Input Modelling / 2D Non-LLM-based Vision Language Action Models