Unverified50% confidenceFactExact time
Typical representatives of action models are vision-language-action models (VLA), such as RT-2 and OpenVLA, which usually use a pretrained multimodal large model as the backbone and, after fine-tuning on robot data, directly output joint positions, end-effector trajectories or discrete action tokens.
1
Sources
50%
Confidence
Long-term
Relevance
10/13/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedVLA (Vision-Language-Action) policy is a paradigm in robotics that unifies visual perception, natural language understanding, and robot action control within a single model framework77% similarUnverifiedVLA将机器人的物理动作表示为动作token，在统一序列预测框架中端到端完成从感知到行动的推理，Google DeepMind的RT-2系列工作是这一范式的代表77% similarUnverifiedCommonly used technical approaches in the industry for controlling robotic arms with large models include direct prompt engineering, Code-as-Action, and specially trained vision-action models (such as RT-2 and π0)72% similarUnverifiedDify、Coze和OpenAI的Agent架构基于ReAct（Reasoning + Acting）框架，模型在每一步先推理再行动再观察67% similarUnverifiedThe capabilities of robots on-site at WRC 2026 are mostly driven by VLA (Vision-Language-Action models) or World Models.67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1022357API
curl https://kongchang.com/api/v1/knowledge/claims/1022357MCP
get_claim(id=1022357)