Abstract
Vision-Language-Action (VLA) models are commonly treated as end-to-end action policies conditioned on natural-language task descriptions. However, their behavior often depends sharply on how the instruction is phrased, suggesting that language is not merely a task label but an optimizable conditioning input. We study whether frozen VLA policies can be improved by optimizing language space rather than updating action weights.
Our method introduces a language-conditioning space policy that translates a human instruction into a short VLA-grounded command using object appearance, spatial relations, and target-grounding cues. The language-conditioning space policy is initialized with a failure-derived command-space prior and optimized with reinforcement learning from sparse task-completion rewards, while the downstream VLA remains fully frozen. This yields language-conditioning space optimization: RL discovers which VLA-grounded commands best elicit successful behavior from the frozen action policy.
Experiments on RL4VLA and VL-Think show that language-conditioning space optimization improves success on instruction-sensitive, symbolic, and multi-object manipulation tasks, demonstrating that language can serve as an optimizable variable for robot foundation models.
Blogger's Review: This research showcases how enhancing VLA models through language input optimization rather than model weight adjustment can significantly improve performance. It offers an innovative perspective, emphasizing the importance and potential of language in robotic learning. Future studies could explore more complex language conditioning spaces and their adaptability to various tasks.