Paper page - Dynin-Robotics: Omnimodal Unified Diffusion Vision-Language-Action Model

Source: https://huggingface.co/papers/2609.13053

Abstract

A shared trajectory model unifies goal and dynamics prediction with action generation for language-conditioned robot control, improving adaptation and success through joint denoising and test-time scaling.

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 sharedtrajectory model. Dynin-Robotics implements this formulation on Dynin-Omni, anomnimodal masked-diffusionbackbone, representing language, visual observations, goals, and actions asdiscrete tokens. By varying conditioning and target spans, the same model learns action prediction,action-conditioned next-observation prediction,terminal goal-state prediction, andtrajectory-to-instruction reconstruction. These interfaces supporttest-time scalingthrough 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 48Open X-Embodimentdatasets 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 withjoint action-next-state denoisingfurther 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 sharedtrajectory modeling as a common interface for learning complementary robot objectives and composing their predictions during control.

View arXiv pageView PDFProject pageGitHub1Add to collection

Get this paper in your agent:

hf papers read 2609\.13053

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper1

#### snu-aidas/Dynin-Robotics

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2609.13053 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2609.13053 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.