Public KG record
Dyna-2
Recorded overview
One generative model that denoises future video and a future action chunk, built
An evidence-backed entity from the project knowledge graph; detailed tool metadata is still being added.
Recorded details
- Benchmarks: xdof ABC
- Capabilities: 预测未来视频或世界状态
- Capabilities: 生成机器人可执行的物理动作
- Capabilities: 预测未来视频和世界未来状态
- Capabilities: 生成机器人可在物理世界执行的动作
- Capabilities: Cross-embodiment generalization
- Capabilities: DiT stack
- Supported types: 接收机器人摄像头采集的视频或图像作为输入
Evidence coverage
The KG currently records 8 known information dimensions and has 10 undocumented dimensions, linked to 2 public source articles. These counts describe record coverage, not model performance or verified fit.
Related public records
- Dyna Robotics — develops
- Action loss — part of
- Flow matching — related to
- Power law — applied to
- Video loss — part of