Elastic Queries Reinforcement Learning: Self-Aware Policy Execution for VLA Models
About
Vision-language-action (VLA) models are powerful action generators for robot manipulation, but they are typically executed with fixed inference and replanning schedules. This rigidity ignores the uneven difficulty of robot control: contact-rich or uncertain states may need more computation and fresher feedback, while easier states can often be handled with fewer inference steps and longer open-loop execution. We propose Elastic Queries Reinforcement Learning (EQRL), a framework that makes each VLA policy query elastic. A lightweight latent-schedule adaptor jointly selects the latent input, denoising budget, and action chunk length, without fine-tuning the underlying VLA model. To make scheduling difficulty-aware, EQRL trains a critic over the joint latent-schedule action and derives a state difficulty signal from critic ensemble disagreement. This signal guides compute toward difficult states, while a learned residual allows task-driven correction. We formulate variable chunk execution as query-level macro-action RL with chunk-dependent discounting and an amortized number-of-function-evaluations (NFE) budget. Across simulation and real-robot manipulation, EQRL reduces amortized inference cost while preserving or improving task success.
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Robot Manipulation | LIBERO 4Tasks | Final Success Rate96 | 4 | |
| Robot Manipulation | ALOHA-Cube | Final Success Rate100 | 4 | |
| Average | Real-robot Offline | Success Rate83.8 | 2 | |
| Pour | Real-robot Offline | Success Rate75 | 2 | |
| Robotic Manipulation | Real-robot Online (Evaluation) | Success Rate95 | 2 | |
| Cucumber | Real-robot Offline | Success Rate90 | 2 | |
| Fruit | Real-robot Offline | Success Rate85 | 2 | |
| Handover | Real-robot Offline | Success Rate85 | 2 |