arXiv AI By Suvarna Kadam (Witsense AI), Soran Parsa (University of Huddersfield)

Vision-Language-Action Models: A Life-Cycle Survey of Design, Training and Deployment

Read the original on arXiv AI →

Vision‑Language‑Action (VLA) models translate web‑pretrained vision‑language knowledge into robot actions, enabling physical control from online data. This survey examines 88 VLA studies from Dec 2022 to Oct 2026, covering manipulation, navigation, humanoid control, and world‑action models, and evaluates them across seven life‑cycle axes: inputs, architecture, action output, intermediate predictions, pretraining data, post‑training adaptation, and real‑time deployment. Key trends include expanding training data, a split in model sizes (3–7 B vs. <1 B parameters), and the rise of generalist VLAs that handle both manipulation and navigation. "whyItMatters":"The survey highlights how VLA models are evolving to leverage broader data sources and more efficient architectures, indicating a shift toward scalable, generalist robotic systems that can be trained on consumer hardware and deployed in real‑world settings."

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv AI.

arXiv Computer Vision
Aug 31

Beyond Data Scaling: Representation-Centric Continued Pre-training for Vision-Language-Action Models

The paper introduces VLAct, a Vision‑Language‑Action model that focuses on representation‑centric continued pre‑training rather than merely scaling robot data. VLAct is trained on diverse, multi‑embodiment robot data and preserves a broad VLM prior while encouraging shared action semantics across embodiments. Experiments across simulation, real‑world, and unseen‑embodiment settings show that VLAct consistently outperforms existing industrial VLA systems, achieving high success rates with only a modest compute budget and open‑source data.

By Senqiao Yang, Chengyao Wang, Yuxin Chen, Zixuan Wang, Longxiang Tang, Haokun Gui, Jinhui Ye, Changsheng Lu, Xiaoyang Wu, Mingkang Zhu, Pengguang Chen, Shu Liu, Zhuotao Tian, Hengshuang Zhao, Bei Yu, Jiaya Jia
arXiv Computer Vision
Oct 28

PDRVLA: A Vision-Language-Action Model with Physical Dynamics Representations

PDRVLA is a Vision‑Language‑Action model that incorporates a lightweight Physical Dynamics Module to extract physical tokens from vision‑language features, enabling action generation without online video synthesis. The module predicts future RGB, depth, and optical‑flow features, guided by auxiliary supervision from a frozen pretrained video world model. Trained solely on publicly available data augmented with human‑to‑robot synthesis, subtask annotation, and embodied chain‑of‑thought (ECoT) annotation, PDRVLA achieves high success rates across four simulation benchmarks, with ablations demonstrating the benefits of world‑model supervision and various conditioning strategies.

By Hanqing Wang, Zhaotian Cai, Ruichuan An, Bohan Zeng, Tao Zhang, Yilin Chen, Xu Jiang, Kaiquan Chen, Chongyang Wu, Haoran Li, Fan Yang, Mingan Lin, Wentao Zhang, Zenan Zhou, Youzheng Wu
arXiv AI
Aug 13

G0.5: One Autoregressive Stream for Robot Reasoning and Action

arXiv:2608. 11739v1 Announce Type: cross Abstract: The prevailing recipe for Vision-Language-Action (VLA) models couples a pretrained VLM with a separately trained flow-matching action expert.

By Yicheng Liu, Zibin Dong, Baijun Ye, Tianyuan Yuan, Tao Jiang, Anqi Yang, Shicheng Cao, Haonan Liu, Yue Sun, Zihan Guo, Xiao Liu, Dong Ke, Changxun Pan, Chenru Wu, Tailai Cheng, Xiaoshu Ren, Xinlei Zhang, Jianning Cui, Zijie Zhao, Haoyu Zhang, Kaiming Xu, Haodong Yang, Bowen Zhang, Jiahui Niu, Shaoting Zhu, Shiduo Zhang, Hang Zhao
arXiv Machine Learning
Nov 6

La-JEPA: Compact Robot Policies with Latent World Modeling

La‑JEPA introduces a compact world‑action model that trains a policy transformer from scratch in the feature space of a pretrained V‑JEPA encoder, learning to predict future features jointly with actions during training but generating only actions at inference. The approach achieves state‑of‑the‑art performance on RoboTwin 2.0 and LIBERO‑Plus while using far fewer parameters and running faster than comparable methods such as Fast‑WAM, LeapBot‑WA, and JEPA‑WAM. A design study shows that JEPA features outperform video VAE latents, especially out‑of‑distribution, and that training‑time future‑prediction choices significantly influence policy performance even when future prediction is omitted at inference.

By Mateo Cl\'emente, George Z. Xue, Amin Abyaneh, Liheng Ma, Rui Heng Yang, Tongtong Cao, Zhanguang Zhang, Yingxue Zhang