Links
Abstract
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. Experiments on simulated and real-robot manipulation tasks demonstrate that AR-VLA can effectively replace traditional chunk-based action heads for both specialist and generalist policies, with superior history awareness and substantially smoother action trajectories while maintaining or exceeding the task success rates of state-of-the-art reactive VLAs.
