Self-Improving VLA Policies: Selected Diffusion Noise for Spurious-Robust Action Smoothing

2026-06-11

AI Summary

Diffusion-based vision-language-action (VLA) policies are susceptible to spurious visual correlations and action noise, leading to fragile performance under perturbations. This work proposes Selected Diffusion Noise (SDN), a training-free, test-time method that treats diffusion noise as a controllable degree of freedom for the first time. SDN dynamically selects the noise vector that maximizes divergence from a reference set while jointly optimizing trajectory coherence to achieve robustness and smooth control. Without modifying model parameters, the approach significantly enhances policy robustness—improving task success rates by 8% in simulation and 10% on real robotic platforms—and effectively suppresses action jitter, thereby increasing behavioral stability.

Abstract

Diffusion-based Vision-Language-Action (VLA) policies enable strong generalization in robotic manipulation, but remain sensitive to spurious visual correlations and noisy action generation, leading to brittle behavior under perturbations. We introduce Selected Diffusion Noise (SDN), a simple, training-free test-time method that improves both robustness and success rate by leveraging the diffusion noise space as a controllable degree of freedom. SDN dynamically samples noise vectors that are maximally separated from a reference set to mitigate reliance on spurious cues, while selecting candidates that yield more coherent action trajectories. This dual objective encourages stable behavior even under object-masked observations and reduces action jitter without modifying model parameters. We evaluate SDN on two simulation benchmarks (Google Robot, Widow-X) and two real-world robotic datasets across multiple VLA policies, including pi_0, Groot-N1.5, and Groot-N1.6. SDN consistently improves success rates by +8% in simulation and +10% in real-world settings, while producing smoother and more stable actions. Our results highlight that diffusion noise selection can serve as an effective and general mechanism for enhancing VLA policies at test time.

Problem

• Vision-Language-Action
• spurious correlations
• action smoothing
• diffusion noise
• robustness

Innovation

• Selected Diffusion Noise
• Vision-Language-Action
• test-time robustness
• action smoothing
• spurious correlation