Pith. sign in

REVIEW 10 cited by

Dita: Scaling Diffusion Transformer for Generalist Vision-Language-Action Policy

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2503.19757 v2 pith:PFUQQ4JE submitted 2025-03-25 cs.RO cs.CV

Zhi Hou , Tianyi Zhang , Yuwen Xiong , Haonan Duan , Hengjun Pu , Ronglei Tong , Chengyang Zhao , Xizhou Zhu
show 3 more authors
Yu Qiao Jifeng Dai Yuntao Chen
This is my paper
classification cs.ROcs.CV
keywords actionditadiffusiontaskstransformeracrossactionscamera
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

While recent vision-language-action models trained on diverse robot datasets exhibit promising generalization capabilities with limited in-domain data, their reliance on compact action heads to predict discretized or continuous actions constrains adaptability to heterogeneous action spaces. We present Dita, a scalable framework that leverages Transformer architectures to directly denoise continuous action sequences through a unified multimodal diffusion process. Departing from prior methods that condition denoising on fused embeddings via shallow networks, Dita employs in-context conditioning -- enabling fine-grained alignment between denoised actions and raw visual tokens from historical observations. This design explicitly models action deltas and environmental nuances. By scaling the diffusion action denoiser alongside the Transformer's scalability, Dita effectively integrates cross-embodiment datasets across diverse camera perspectives, observation scenes, tasks, and action spaces. Such synergy enhances robustness against various variances and facilitates the successful execution of long-horizon tasks. Evaluations across extensive benchmarks demonstrate state-of-the-art or comparative performance in simulation. Notably, Dita achieves robust real-world adaptation to environmental variances and complex long-horizon tasks through 10-shot finetuning, using only third-person camera inputs. The architecture establishes a versatile, lightweight and open-source baseline for generalist robot policy learning. Project Page: https://robodita.github.io.

Discussion (0). Sign in to comment.

Forward citations

Cited by 10 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Nautilus: From One Prompt to Plug-and-Play Robot Learning

    cs.RO 2026-05 unverdicted novelty 6.0 of 10

    NAUTILUS is a prompt-driven harness that automates plug-and-play adapters, typed contracts, and validation for policies, benchmarks, and robots in learning research.

  2. Think Proprioceptively: State-Grounded Visual Token Selection for VLA Policies

    cs.RO 2026-02 conditional novelty 6.0 of 10

    Using tokenized proprioception plus instruction to select ~15% of visual patches matches or beats full-token VLA baselines and cuts latency by ~58%.

  3. CogVLA: Cognition-Aligned Vision-Language-Action Model via Instruction-Driven Routing & Sparsification

    cs.CV 2025-08 conditional novelty 6.0 of 10

    CogVLA pairs instruction-conditioned visual-token aggregation (EFA-Routing) with transformer-layer pruning (LFP-Routing) and bidirectional action decoding (CAtten), reporting LIBERO 97.4%, real-world 70.0%, 2.5x less ...

  4. GeoVLA: Empowering 3D Representations in Vision-Language-Action Models

    cs.RO 2025-08 conditional novelty 6.0 of 10

    A robot policy that combines 2D vision-language features with a point-cloud encoder and a mixture-of-experts diffusion action head reports SOTA manipulation success in simulation and robust real-world behavior under h...

  5. Visual Embodied Brain: Let Multimodal Large Language Models See, Think, and Control in Spaces

    cs.CV 2025-05 reject novelty 6.0 of 10

    VeBrain unifies perception, spatial reasoning, and robot control in one MLLM by representing control as keypoint detection plus skill selection, with a robotic adapter for deployment.

  6. Balancing Signal and Variance: Adaptive Offline RL Post-Training for VLA Flow Models

    cs.RO 2025-09 conditional novelty 5.0 of 10

    ARFM adaptively adjusts a scaling factor in the flow-matching loss so that offline RL advantage signals are preserved while gradient variance is controlled, improving VLA robot policy fine-tuning.

  7. Align-Then-stEer: Adapting the Vision-Language Action Models through Unified Latent Guidance

    cs.RO 2025-09 conditional novelty 5.0 of 10

    A plug-and-play fine-tuning method using two VAEs and a latent-distance guidance loss improves cross-embodiment and cross-task success rates of diffusion- and flow-based VLA policies.

  8. Source Component Shift Adaptation via Offline Decomposition and Online Mixing Approach

    cs.LG 2025-08 unverdicted novelty 5.0 of 10

    The claimed result is that source-component-shift adaptation splits cleanly into offline component learning via EM and online mixing-weight updates, cutting cumulative test loss by up to 67.4%.

  9. StemVLA:An Open-Source Vision-Language-Action Model with Future 3D Spatial Geometry Knowledge and 4D Historical Representation

    cs.RO 2026-02 reject novelty 4.0 of 10

    StemVLA supervises a GPT-2-based VLA with predicted future 3D-geometry features (VGGT) and temporally aggregated history, reporting 86.0% on LIBERO-Long - but its CALVIN results and equations are placeholders.

  10. Leveraging OS-Level Primitives for Robotic Action Management

    cs.OS 2025-08 conditional novelty 4.0 of 10

    Applying OS-style exception handling, context caching, and replay to robotic action slices raises success rates 7x to 24x and cuts execution steps up to 74% for repetitive manipulation tasks, without retraining the VLA model.

Pith tools