Drift is a Sampling Error: SNR-Aware Power Distributions for Long-Horizon Robotic Planning
For embodied AI researchers, this work provides a novel inference-time solution to instruction drift in long-horizon tasks, though the improvement is incremental as it builds on existing sampling and search techniques.
This paper redefines instruction drift in long-horizon robotic tasks as a systematic sampling error and proposes Context-Aware Power Sampling (CAPS), a training-free inference-time method that uses power distributions and SNR-based metacognitive control to improve robustness. CAPS achieves substantial improvements over OpenVLA and TACO on RoboTwin, Simpler-WindowX, and Libero-long benchmarks without parameter updates.
Despite rapid progress in Vision-Language-Action (VLA) models for robotic control, instruction drift remains a persistent failure mode in long-horizon tasks. This paper reconceptualizes this phenomenon, positing that instruction drift is fundamentally a systematic sampling error: local greedy sampling is prone to collapsing into "Negative Pivotal Windows"--irreversible local optima with high local probability that sever global success pathways. To address this, we propose Context-Aware Power Sampling (CAPS), a training-free inference-time computation framework. CAPS leverages power distributions to sharpen global trajectory probabilities, enabling lookahead search over the model's conditional generative trajectory distribution. Furthermore, we introduce a metacognitive control mechanism based on Signal-to-Noise Ratio (SNR). This mechanism triggers adaptive MCMC search solely when drift risk is detected, enabling a dynamic transition from "intuitive fast thinking" to "rational slow search." Experiments on RoboTwin, Simpler-WindowX, and Libero-long benchmarks show that CAPS achieves substantial improvements over strong baselines, including OpenVLA and TACO, without parameter updates. These results support the effectiveness of adaptive inference-time computation for improving long-horizon robustness in embodied control.