RLDX-1 robotic policy enhances dexterous manipulation with new transformer architecture

Researchers have introduced RLDX-1, a new robotic policy designed for dexterous manipulation that integrates heterogeneous modalities through a Multi-Stream Action Transformer architecture. This approach aims to overcome limitations in current Vision-Language-Action models by incorporating motion awareness, memory-based decision-making, and physical sensing. RLDX-1 demonstrates superior performance compared to existing models like $\pi_{0.5}$ and GR00T N1.6, particularly in complex real-world tasks and humanoid robot control. AI

Summary written by gemini-2.5-flash-lite from 1 source. How we write summaries →

IMPACT Introduces a novel architecture for dexterous robotic manipulation, potentially advancing capabilities in real-world human-robot interaction.

RANK_REASON This is a technical report detailing a new robotic policy and architecture published on arXiv. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• RLDX-1
• Multi-Stream Action Transformer
• MSAT
• Vision-Language-Action models
• VLAs
• $\pi_{0.5}$
• GR00T N1.6
• ALLEX

• paper
• model release