2025/12/05 by Kevin Black, Black, Kevin, Allen Z. Ren +5 · 1 citation
Computer Science · Engineering · #Artificial Intelligence (cs.AI) #FOS: Computer and information sciences #Multimodal Machine Learning Applications #Reinforcement Learning in Robotics #Robot Manipulation and Learning #Robotics (cs.RO)
paper · pdf · doi:10.48550/arxiv.2512.05964
openalex publication_date 2025/12/05 · openalex created_date 2025/12/09 · openalex updated_date 2026/07/28
Real-time chunking (RTC) enables vision-language-action models (VLAs) to generate smooth, reactive robot trajectories by asynchronously predicting action chunks and conditioning on previously committed actions via inference-time inpainting. However, this inpainting method introduces computational overhead that increases inference latency. In this work, we propose a simple alternative: simulating inference delay at training time and conditioning on action prefixes directly, eliminating any inference-time overhead. Our method requires no modifications to the model architecture or robot runtime, and can be implemented with only a few additional lines of code. In simulated experiments, we find that training-time RTC outperforms inference-time RTC at higher inference delays. In real-world experiments on box building and espresso making tasks with the π0.6 VLA, we demonstrate that training-time RTC maintains both task performance and speed parity with inference-time RTC while being computationally cheaper. Our results suggest that training-time action conditioning is a practical drop-in replacement for inference-time inpainting in real-time robot control.