RA-VLA: Retrieval-Augmented VLA for Test-Time Adaptation

2026-08-26

AI Summary

为解决VLA模型在新任务中表现脆弱的问题，提出RA-VLA框架，通过结合行为一致的上下文检索与执行流程，实现高效的任务适应。

Abstract

Vision-Language-Action (VLA) models provide a versatile foundation for general robotic manipulation, yet they exhibit significant brittleness when confronted with novel task distributions. While In-Context Imitation Learning (ICIL) offers a training-free alternative, existing frameworks suffer from an adaptation bottleneck that hinders the effective translation of expert context to executable actions. This failure originates from superficial retrieval mechanisms and an inherent behavioral inertia that anchors the policy to its pre-trained priors. To address these limitations, we present RA-VLA, a retrieval-augmented VLA framework that integrates behavior-aligned context retrieval with a grounded execution pipeline. By enforcing faithful adherence to functional cues within a scalable architecture, RA-VLA facilitates seamless task adaptation while preserving inference efficiency. Our empirical evaluations across the LIBERO benchmark and a real-world UR5e environment demonstrate that RA-VLA achieves superior success rates and computational efficiency, establishing a robust framework for training-free robotic adaptation.

Problem

• VLA models
• novel task distributions
• In-Context Imitation Learning
• adaptation bottleneck
• behavioral inertia

Innovation

• retrieval-augmented
• behavior-aligned context retrieval
• grounded execution pipeline
• task adaptation