About this paper
BiCAT shares a motion representation across arms and conditions each arm's tokens on its partner via symmetric cross-arm attention and a joint decoder, enabling parallel arm token prediction and improving bimanual manipulation success.
Autoregressive vision-language-action (VLA) models predict discrete action tokens with the next-token objective of their vision-language backbone, so the action tokenizer defines the interface between the policy and the robot. In bimanual manipulation, the two arms share a repertoire of movements yet coordinate through complementary roles, and an action representation should capture both. However, most existing tokenizers encode both arms as a single concatenated action vector, in which motion patterns common to the two arms are not shared and cross-arm dependencies remain entangled within each token. The policy must therefore learn both aspects implicitly. We propose Bimanual Coordination…
autoregressive vision-language-action (VLA) modelsdiscrete action tokensnext-token objectivevision-language backboneaction tokenizerbimanual manipulationaction representationconcatenated action vectorcross-arm dependenciesBimanual Coordination Action Tokenization (BiCAT)
- Upvotes
- 2
- Authors
- 4
- Published
- Nov 17, 2026
Summary and abstract excerpt via Hugging Face Papers ↗ · arXiv.
Trending history
BiCAT: Bimanual Coordination Action Tokenization was first seen on Hugging Face · Daily papers on Nov 19, 2026. Since then it has appeared in 3 checks on 1 different day. Its best position was #25 of 30, reached on Nov 19, 13:29 UTC. It dropped off the list after Nov 19, 15:29 UTC.