Concertina: Embodying Manipulation and Cognition in One Diffusion Vision-Language Model
Overview
This paper presents Concertina, an Embodied Vision-Language Model (eVLM) that unifies manipulation and cognition within a single diffusion backbone. The model encodes control timesteps as continuous latent action tokens and reconstructs temporally ordered action blocks using the same operator as multimodal pretraining. With only 0.21% of parameters dedicated to active action projections, Concertina achieves 98.15% average success on LIBERO, 85.9% on LIBERO-Plus (10,030 tasks), and an average chain length of 4.142 on CALVIN ABC-D. The authors validate cross-time trajectory coupling in the shared backbone layers via an Action Interface Audit. Code and models are available on GitHub and Hugging Face.
Generated from attributed reports · Updated 2 days ago
Event evidence and corrections
0 attributed source owners. Ownership does not establish independent confirmation. Quantities are reported separately and are never added together.
Reported quantity · average completed chain length on CALVIN ABC-D: 4.142 other · Basis not reported
“5.1% dedicated action-expert capacity in representative continuous VLAs. Using our proposed Action Interface Audit, we confirm that cross-time trajectory coupling emerges in the middle and late layers of the shared backbone rather than at the action interfaces. On standard benchmarks, Concertina achieves 98.15% average success on regular LIBERO, 85.9% on the 10,030-task LIBERO-Plus benchmark, and 4.142 average comple”
Exact source · revision 1Source owner not reported
Artifact availability · code: available
“ted chain length on CALVIN ABC-D. Code: https://github.com/lose4578/Concertina Hugging Face: https://huggingface.co/collections/Traly/concertina”
Exact source · revision 1Source owner not reported
Report timeline
Follow attributed reports and material updates.
- arXiv Robotics — research abstractsSignalConcertina: Embodying Manipulation and Cognition in One Diffusion Vision-Language Model
This paper presents Concertina, an Embodied Vision-Language Model (eVLM) that unifies manipulation and cognition within a single diffusion backbone. The model encodes control timesteps as continuous latent action tokens and reconstructs temporally ordered action blocks using the same operator as multimodal pretraining. With only 0.21% of parameters dedicated to active action projections, Concertina achieves 98.15% average success on LIBERO, 85.9% on LIBERO-Plus (10,030 tasks), and an average chain length of 4.142 on CALVIN ABC-D. The authors validate cross-time trajectory coupling in the shared backbone layers via an Action Interface Audit. Code and models are available on GitHub and Hugging Face.
Timezone · UTC
Article dates follow your selected timezone. Briefing editions use Hong Kong time (UTC+8).