AI & ML
Test-Time Improvement of VLA Policies via Selected Diffusion Noise for Spurious-Robust Action Smoothing
arXiv cs.RO0 views
arXiv:2606.14084v2 Announce Type: replace Abstract: Diffusion-based Vision-Language-Action (VLA) policies have demonstrated strong generalization in robot manipulation, but their deployment can remain unreliable due to spurious visual reliance and stochastic action generation, leading to grounding failures and kinematically unstable behaviors. We introduce Selected Diffusion Noise (SDN), a training-free method that treats diffusion noise as a controllable degree of freedom for improving test-time reliability. SDN samples multiple candidate action chunks and applies two-stage selection: it first filters candidates based on task-relevant visual grounding using a counterfactual negative set, then selects the kinematically most stable chunk based on JerkRMS. Experiments on three VLA policies across both simulation and real-world environments show improvements of 8-10% and 18.33%, respectively, demonstrating SDN's effectiveness for test-time improvement.
Source：https://arxiv.org/abs/2606.14084PreviousBridging the Morphology Gap: Adapting VLA Models to Dexterous Manipulation via Intent-Conditioned Fine-TuningNextAutoDex: An Automated Real-World System for Dexterous Grasping Data Collection
Related posts
RegimeVGGT: Layer-Wise Spatially Preserving Redundancy Removal for Visual Geometry Grounded Transformer
0arXiv cs.RO
RoboTryOn: Synthesizing Dexterous Robot Demonstrations from Human Videos via Virtual Try-On
0arXiv cs.RO
Future Representation Distillation for Imagination-free World Action Models
0arXiv cs.RO
When Robots Outweigh Objects: How Inertia Mismatch Shapes Robot Policy Learning
0arXiv cs.RO
Learning to Live: A Step Toward Lifelong Embodied Intelligence
0arXiv cs.RO
Tactus: Tactile & Pose Sensing Glove for Natural Dexterity Data Collection
0arXiv cs.RO