Metraipolis › archive › dc_036fb718fa5b
Dynin-Robotics: Omnimodal Unified Diffusion Vision-Language-Action Model — arXiv:2609.13053
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.13053
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-16T03:03:47.093Z
- SHA-256 of the archived text
342272569da75821cf3df7860996b8460c98edc36093d45a2c4c11ce8ab7eefa- Stable URL of this page
https://metraipolis.com/documento/dc_036fb718fa5b
How to cite
Dynin-Robotics: Omnimodal Unified Diffusion Vision-Language-Action Model — arXiv:2609.13053 — arXiv (en), https://arxiv.org/abs/2609.13053 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_036fb718fa5b (Metraipolis archive dc_036fb718fa5b, sha256 342272569da75821cf3df7860996b8460c98edc36093d45a2c4c11ce8ab7eefa)
BibTeX
@misc{dc036fb718fa5b,
title = {Dynin-Robotics: Omnimodal Unified Diffusion Vision-Language-Action Model — arXiv:2609.13053},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.13053},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_036fb718fa5b; sha256 342272569da75821cf3df7860996b8460c98edc36093d45a2c4c11ce8ab7eefa},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.13053. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Visual goal and dynamics prediction can provide language-conditioned robot policies with both a target outcome and a representation of action-dependent scene changes. We bring these predictions into action generation and selection through a shared trajectory model. Dynin-Robotics implements this formulation on Dynin-Omni, an omnimodal masked-diffusion backbone, representing language, visual observations, goals, and actions as discrete tokens. By varying conditioning and target spans, the same model learns action prediction, action-conditioned next-observation prediction, terminal goal-state prediction, and trajectory-to-instruction reconstruction. These interfaces support test-time scaling through goal prediction, action-candidate evaluation, and joint refinement of action and future-state predictions. We continually pretrain the model on approximately 1.33 million trajectories from 48 Open X-Embodiment datasets and adapt it separately to downstream domains. On two VLABench tasks, robot pretraining improves adaptation within a fixed Stage-2 step budget, and the full objective mixture improves shifted-instruction success over Policy-only post-training under the same coupled decoder. Combining goal guidance with joint action-next-state denoising further improves shifted-instruction success over action-only decoding; the benefit depends on how the predictions are composed. Dynin-Robotics achieves competitive performance on LIBERO and zero-shot LIBERO-Plus, together with a 78.4% average success rate across four manipulation conditions on a Franka Research 3 robot. An optimized block-parallel implementation accelerates model-side action decoding by up to 29.2x relative to the base implementation under the reported profiling setup. These results support shared trajectory modeling as a common interface for learning complementary robot objectives and composing their predictions during control. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-14 in cs.AI). Categorie arXiv: Robotics (cs.RO) ; Artificial Intelligence (cs.AI); Machine Learning (cs.LG). Pagina del lavoro: https://arxiv.org/abs/2609.13053 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-16T02:51:27.975Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-14?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).