The authors assembled 1,695,684 cleaned training anchors from multiple sources and trained Robota-Embed to combine language instructions, visual observations, and native robot actions.
Action-aware embeddings retrieve better robot experiences for future actions
Robota-Embed matches a robot’s current task and execution stage to useful recorded continuations, improving retrieval and downstream policy performance.
Big Tech
Yunze Liu · Jing Wang · Shenrui Wu · Wenkun He · Guanqi Zhan · Xianzheng Ma · +13 more
NVIDIA · MIT · UCB · Oxford · CMU
Research Digest··3 min read
Liu and colleagues trained a vision-language-action embedding model that represents both a robot’s present state and the future continuation of prior experience as searchable vectors.
Why this paper
From NVIDIA and 4 others
In one line
Robota-Embed jointly encodes vision, language, and action for robot experience retrieval, outperforming action-free embeddings.
What we could check
- ·No code link found
- ·No weights link found
- ·No dataset link found
- ·No compute details found
- ·No stated limitations found
- ✓Reports numbers on named benchmarks (5 benchmarks)
Observed from the paper text and links we have. Absence here means we did not find it, not that it does not exist.
§