Unverified50% confidenceFactExact time
Google DeepMind 的 RT-2 和 Figure AI 等机器人项目为收集人类行为数据投入了巨额成本
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
7/19/2026
First Seen
Valid until: 10/17/2026
Sources
Related Claims
UnverifiedGoogle DeepMind于2023年发布的RT-2首次验证了在550亿参数的视觉语言模型上微调机器人控制的可行性,其泛化能力相比前代提升近一倍71% similarUnverifiedGoogle DeepMind推动了A2A(Agent-to-Agent)通信规范69% similarUnverifiedGoogle的RT-2、Open X-Embodiment等项目证明跨机器人平台的大规模数据集能显著提升策略泛化能力68% similarUnverifiedGoogle DeepMind的RT-2(2023年)将机器人动作词元化为离散token,附加到视觉-语言模型PaLI-X和PaLM-E的输出词汇表中,展现出零样本泛化能力68% similarUnverified学术界正探索基础模型在机器人领域的应用,如Google DeepMind的RT-2和斯坦福的Open X-Embodiment项目68% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/558980API
curl https://kongchang.com/api/v1/knowledge/claims/558980MCP
get_claim(id=558980)