AR-VLA: True Autoregressive Action Expert for Vision-Language-Action Models
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866916000601997312 |
|---|---|
| author | Hu, Yutong Zaech, Jan-Nico Nikolov, Nikolay Yao, Yuanqi Dey, Sombit Albanese, Giuliano Detry, Renaud Van Gool, Luc Paudel, Danda |
| author_facet | Hu, Yutong Zaech, Jan-Nico Nikolov, Nikolay Yao, Yuanqi Dey, Sombit Albanese, Giuliano Detry, Renaud Van Gool, Luc Paudel, Danda |
| contents | We propose a standalone autoregressive (AR) Action Expert that generates actions as a continuous causal sequence while conditioning on refreshable vision-language prefixes. In contrast to existing Vision-Language-Action (VLA) models and diffusion policies that reset temporal context with each new observation and predict actions reactively, our Action Expert maintains its own history through a long-lived memory and is inherently context-aware. This structure addresses the frequency mismatch between fast control and slow reasoning, enabling efficient independent pretraining of kinematic syntax and modular integration with heavy perception backbones, naturally ensuring spatio-temporally consistent action generation across frames. To synchronize these asynchronous hybrid V-L-A modalities, we utilize a re-anchoring mechanism that mathematically accounts for perception staleness during both training and inference. Experiments on simulated and real-robot manipulation tasks demonstrate that the proposed method can effectively replace traditional chunk-based action heads for both specialist and generalist policies. AR-VLA exhibits superior history awareness and substantially smoother action trajectories while maintaining or exceeding the task success rates of state-of-the-art reactive VLAs. Overall, our work introduces a scalable, context-aware action generation schema that provides a robust structural foundation for training effective robotic policies. Code and Videos available at https://arvla.insait.ai |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_10126 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | AR-VLA: True Autoregressive Action Expert for Vision-Language-Action Models Hu, Yutong Zaech, Jan-Nico Nikolov, Nikolay Yao, Yuanqi Dey, Sombit Albanese, Giuliano Detry, Renaud Van Gool, Luc Paudel, Danda Robotics Artificial Intelligence We propose a standalone autoregressive (AR) Action Expert that generates actions as a continuous causal sequence while conditioning on refreshable vision-language prefixes. In contrast to existing Vision-Language-Action (VLA) models and diffusion policies that reset temporal context with each new observation and predict actions reactively, our Action Expert maintains its own history through a long-lived memory and is inherently context-aware. This structure addresses the frequency mismatch between fast control and slow reasoning, enabling efficient independent pretraining of kinematic syntax and modular integration with heavy perception backbones, naturally ensuring spatio-temporally consistent action generation across frames. To synchronize these asynchronous hybrid V-L-A modalities, we utilize a re-anchoring mechanism that mathematically accounts for perception staleness during both training and inference. Experiments on simulated and real-robot manipulation tasks demonstrate that the proposed method can effectively replace traditional chunk-based action heads for both specialist and generalist policies. AR-VLA exhibits superior history awareness and substantially smoother action trajectories while maintaining or exceeding the task success rates of state-of-the-art reactive VLAs. Overall, our work introduces a scalable, context-aware action generation schema that provides a robust structural foundation for training effective robotic policies. Code and Videos available at https://arvla.insait.ai |
| title | AR-VLA: True Autoregressive Action Expert for Vision-Language-Action Models |
| topic | Robotics Artificial Intelligence |
| url | https://arxiv.org/abs/2603.10126 |