Ant Group Robbyant Launches LingBot-VA 2.0: Native Physical AI Video-Action Model
Decision Brief
What changedAnt Group Robbyant released the LingBot-VA 2.0 technical report, a video-action foundation model built from scratch for physical AI.
Why it mattersWith predictive reasoning and 225 Hz asynchronous control, it offers research teams a more realistic and responsive option for embodied intelligence.
Who should careAll AI builders
Affected stackNo specific stack identified
Source confidenceMedium · Reliable media or first-hand reporting
LingBot-VA 2.0, developed by Ant Group Robbyant, is a physical AI video-action foundation model built from scratch for embodied intelligence, unlike common approaches that fine-tune video generators. Core capabilities include predictive reasoning to anticipate future states before execution, re-grounding after each real observation, and 225 Hz asynchronous control. Technical aspects involve causal DiT (Diffusion Transformer), sparse MoE video streams, and semantic visual-action tokenizers.
Summary basis: official / RSS sourceCompiled from the source scope noted above; the original remains authoritative.
Sources
- MarkTechPost
Fast research-paper and ML tooling summaries, useful for infra and agent updates.
- MarkTechPost
留言
登入后即可留言，和其他 builder 交换实测心得。
还没有留言，抢头香。