ACT³：面向动作的三流Transformer VLA模型

1. 12:00
   arXiv · Robotics

   ACT³：面向动作的三流 Transformer，重构语义、动力学与控制

   研究者提出 ACT³（Action-Centric Tri-Stream Transformer），一种面向动作的三流架构，将 VLM 的语义理解与视频生成 World Model 的动力学预测融合进控制动作，同时保留各上下文流的独立角色。

Current heat 9·Comparable peak 10（Oct 9）·Comparable change over 24 hours –

The trend compares only the same participants observed continuously; its range may be smaller than the current heat count. Move or click on the chart to inspect hourly heat; use the left and right arrow keys to switch.